# fal x Sequoia 72-Hour Video Hackathon: Project Gallery - **Event:** [fal x Sequoia 72-Hour Video Hackathon](https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon) - **When:** Jul 17 at 12:00 PM – Jul 19 at 8:00 PM (PDT) - **Where:** Online - **Hosts:** [fal](https://cerebralvalley.ai/u/fal), [Sequoia Capital](https://cerebralvalley.ai/u/sequoia) - **Projects:** 105 (6 placed) - **Page:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery ## Projects ### 1. Reverie Everything you watch has one look: the one it shipped with. Reverie lets you change it to match your taste, your mood, whatever you feel like today. Point it at any video and it repaints every frame into a new world while it plays: comic book, watercolor, anime, oil painting, film noir. Type a sentence and the whole look shifts as you watch. Under the hood, Reverie is the real-time layer on top of the model, not the model. It runs on fal with Decart Lucy 2.5 over a WebRTC bridge, detects a video's cuts and pre-warms a pool of live sessions so the style hands off cleanly at every hard cut, and re-steers the moment you type a new prompt, all with the key server-side. More of what fills a screen is generated now, not just recorded, and that share only grows. This was never only about movies: recolor your feed like a graphic novel, drop your game into a Ghibli afternoon, turn last night's call into a memory that looks the way it felt. Everything on your screen, finally yours. - **Team:** [dylan merigaud](https://cerebralvalley.ai/u/dylanmerigaud) - **GitHub:** https://github.com/DylanMerigaud/reverie - **Demo video:** https://youtu.be/lhAVPWUvQP0 - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=16 ### 2. J3 Rest, As Discovered begins like a serious scientific inquiry: objects don't fall the way we expect, sunlight follows patterns no one has explained, and a researcher goes looking for the reason why. What he finds isn't first contact, isn't a message, and isn't remotely what anyone was trained to expect: it's something that was already here, long before the investigation started, unbothered and entirely uninterested in being discovered at all. The closer he gets to proving it, the more the search starts to look less like a hunt for answers and more like a quiet confession about everything humans do to avoid sitting still - the schedules, the timelines, the permission we keep trying to earn just to exist. Deadpan and quietly absurd, it's a mockumentary about the biggest discovery in human history, and how it turned out to require absolutely nothing from us at all. - **Team:** [Jingjing Ji](https://cerebralvalley.ai/u/Jingjing) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=17 ### 3. dwlsdigital (solo) BAD WEEK is an animated dark comedy short told from the other side of the apocalypse. Three zombies share lunch at an abandoned food court and swap stories about their terrible week: an awkward encounter with a fleeing human, a hunt ruined by a loyal guard dog (who is now an unwanted pet), and a "perfect" ambush that turned out to be a military trap. Just as they conclude humans aren't worth the trouble — a survivor who has been listening the whole time delivers the film's final review. Produced solo in 72 hours with a cinematography-first pipeline: original character sheets, 9-panel storyboards, and shot-by-shot prompts with locked spatial geography, chest-rig POV camera work, and strict continuity rules. Video generated with Seedance on fal; several action sequences were generated on Higgsfield after repeated 422 moderation errors on safe content (reported in Discord). Voices re-dubbed with ElevenLabs for cross-clip consistency, music via Suno, edited in CapCut. ARCANE-inspired painterly style, built entirely from original character designs. - **Placement:** Finalist - **Team:** [Rizki Nurulloh](https://cerebralvalley.ai/u/dwlsdigital) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=24 ### 4. PrismFlow PrismFlow is the "Cursor for AI Generated Video". It gives AI agents real context and real editing power, but keeps the human in a review loop where every change is a visible, reversible, priced proposal. AI can now generate video clips, but editing with AI is broken in two ways. First, indie creators and solo media makers drown in toil — trimming, nudging, reordering clips, tightening pacing, hunting for "that one shot" by scrubbing through footage. Second, the existing "AI editor" answer is a chatbot bolted next to a timeline is not enough, you need the agent to do more of the busy work for you and give you choice on what to do next. 1. Agents propose diffs, they don't mutate. You tell an editing agent "tighten the pacing in act two"; it works the timeline and produces a changeset — ghost clips rendered semi-transparently on the timeline, which you preview, accept, reject, or rebase per change, exactly like reviewing a code diff. The agent's tool calls are visible in a rail, like Cursor's. 2. The project is indexed like a codebase. Every clip carries provenance (prompt, model, seed, parent asset) plus derived context — a vision-model description of each shot and audio transcripts, embedded locally. That powers semantic search ("the shot where the fox jumps") and gives agents real context about your project, not generic prompts. 3. Clips are cached evaluations of prompts, not files. Because provenance is kept, any clip can be re-rolled, prompt-edited, or regenerated with the same prompt on a different model — with semantic search over the entire FAL model catalog to find a better model, side-by-side variant comparison, and the winner dropped back into the timeline. Around that core: story-first creation (script beats generate scenes; the screenplay and timeline stay bound), character and style locking (cast a character once, every generation is forced through the locked references so it stays coherent across shots), and cost transparency (per-model pricing shown before you generate, half-cost draft tiers for iteration, and a live project spend dashboard). - **Team:** [Anup Ghatage](https://cerebralvalley.ai/u/ghatage) - **GitHub:** https://github.com/Ghatage/PrismFlow - **Demo video:** https://www.youtube.com/watch?v=oFMFrujFdBQ - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=18 ### 5. Action Leopard Action Leopard is spatial control and reasoning for action scenes in generative video. The app, its prompting system, and its reference image creation and segmentation tools are built for shots where location matters and continuity can't be approximated. Video models drift. Cars grow grilles facing the wrong direction, and action reverts to predictable scenarios from the models' training, making it hard to force objects and characters to do unique and surprising things while still following the rules of physics. Action Leopard tackles this in two ways. First, you upload a script and a set of ordered location photos that map a real place. An agentic loop plans shots and generates start frames, then a separate judge model checks each frame against the location, the script, and scene invariants (e.g., which character or vehicle is ahead) and rewrites prompts until physics and continuity hold. Second, when the loop hits something it can't fix, it hands the shot to you. Segmentation cuts characters, vehicles, and props out of your location photo as movable stickers over a clean background plate, so you can rearrange things on a canvas, annotate motion paths, and stage the reference frame yourself. - **Team:** [Cortney Harrington](https://cerebralvalley.ai/u/cineleopard) - **GitHub:** https://github.com/CoreOrca/action-leopard - **Demo video:** https://youtu.be/tepnKMHZ2Jo - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=25 ### 6. MrNineTeam Muc waters a tiny seed every morning, but nothing seems to change. The soil stays still, and he starts to feel lonely, heavy, and unseen. After a quiet rain, one small green leaf finally appears. Muc realizes that growth often happens silently, hidden beneath the surface, before anyone can see it. The message is: slow progress still matters. Feeling heavy does not mean you are failing. Small acts of care, patience, and asking for help when needed can slowly bring hope back. - **Team:** [Nine Mr](https://cerebralvalley.ai/u/mrnine) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=1 ### 7. Fash.studio John, a cave diver, follows a white fish deeper into an unmapped tunnel system, chasing it out of curiosity. Caught up in the thrill of discovery, he loses track of distance until he realizes the passage behind him has narrowed too much to turn back. Wedged in the dark with his oxygen dropping, he finds a small air pocket and pulls himself into it, gasping. In the flickering light, the white fish reappears, hovering just beyond the surface, and begins to speak. Its words don't show him a way out. Instead, they give him something else: a sudden, clear-headed burst of will. Snapping back to the cold reality of the cave, John grabs his oxygen tank and starts striking it against the rock, again and again, the sound echoing through the tunnels. Above ground, a rescue team catches the faint metallic rhythm and traces it back to him. He surfaces gasping into light. A story about how imagination doesn't always show us the way out, but it can give us the strength to find it ourselves. - **Team:** [Sumeyye Zehra Dizlek](https://cerebralvalley.ai/u/Zehra), [Mustafa Erdem](https://cerebralvalley.ai/u/FashStudio) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=3 ### 8. Latent A man sits in an office papered with his past: diplomas, group photos, framed proof of a life in order. One by one, the diplomas and pictures go blank. From a drawer he takes out the one picture that never made it to the wall, the woman he loved. In the movie the viewer follows him into the architecture of his own memory: a hall of filmstrips where the two of them are still dancing, a darkroom where every print of her dissolves in his hands, a maze of mirrors crowded with versions of himself he no longer recognizes. He cannot fix the image. - **Team:** [Sara Rutz](https://cerebralvalley.ai/u/sararutz) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=19 ### 9. Plumlee Productions Plumlee, a kindhearted polar bear, uncovers a dark secret that forces him to question his identity, his reality, and what gives life meaning in a world shaped by AI. - **Team:** [Lauren S](https://cerebralvalley.ai/u/lauren), [Scott Maloff](https://cerebralvalley.ai/u/SuperGenius) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=20 ### 10. Valezart "The Invincible Armada" is a cinematic concept trailer that reimagines one of history's most dramatic turning points. The narrative captures the high-stakes geopolitical chess match between a zealous King Philip II of Spain and an unyielding Queen Elizabeth I of England. As the colossal Spanish fleet of 130 vessels sails to open the English Channel, they are forced into a desperate retreat around the treacherous coasts of Scotland and Ireland. Pushed to the brink by agile English warships, the Armada's true doom comes from an apocalyptic tempest. This trailer tells a story of hubris, faith, and survival, showing how an empire's ambition was ultimately crushed not by men, but by the unforgiving forces of nature. - **Team:** [Albert Alvarez](https://cerebralvalley.ai/u/Valezart) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=26 ### 11. Take 47 A grandchild rebuilds their grandfather's favorite memory: the legendary winter of 1963, when the whole town lived on the frozen canal. The archive says it never happened. He corrects every take anyway. The film is structured as numbered takes — each correction regenerates the memory, and the takes drift the way memory drifts. Generative video's instability becomes the film's grammar of forgetting. Built entirely on fal: Nano Banana 2 for identity-locked first frames, face-checked before any video spend; Veo 3.1 and Kling 3.0 Pro cast per shot after a logged three-model bake-off. The 23-second closing glide is two Kling generations joined by a forced match-frame — a single invisible cut. The grandfather's voice-over is a MiniMax clone of his own AI-generated on-camera performance; an independent transcriber reads them as one speaker. It never happened. It's still true. - **Team:** [kiran Nreddy](https://cerebralvalley.ai/u/Kiranaero) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=28 ### 12. DouNiu Paris, 2145. Monet, the last of the human Impressionist masters, hates all machines—thirty-three years ago, a medical AI deemed his wife Camille “not worth saving,” and he was forced to watch her die of her illness. His new robot bodyguard, Kalos, is assigned to protect him, yet he is always secretly watching Monet paint from a corner of the studio, saying he wants to learn. Amid the cold metal and the shifting play of light and shadow, a story of memory, loss, and light quietly unfolds. - **Team:** [Dou Niu](https://cerebralvalley.ai/u/NiuDou), [Samuel Xiong](https://cerebralvalley.ai/u/imsamuelxiong) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=2 ### 13. AiStudios.ca High-Concept Hit Elevator Pitch: Indiana Jones meets Mad Max in space! An archaeologist abandons a wounded Earth to follow an ancient map to humanity’s untouched Garden of Eden—only to discover that the paradise is actually our ruined original world, and his arrival may lead its forgotten warlords back to the home he left behind. Official Project Summary EXODUS: ZERO is a new original high-concept science-fiction adventure about Adrian, an archaeologist who devotes his life to an ancient map promising a pristine second Earth—a hidden Eden untouched by humanity. Convinced that salvation lies elsewhere, he abandons a damaged home instead of helping repair it. When Adrian finally reaches the promised world, the discovery overturns everything he believes: this is not a second Earth, but humanity’s original home. Five thousand years earlier, its survivors fled to Adrian’s planet, erased the truth, and began again—only to repeat the same cycle of destruction. The first Earth is now a savage wasteland divided by rival factions and war machines. Adrian’s arrival destabilizes that world and may reveal the location of the civilization he left behind. Forced to choose between escape and responsibility, he evolves from a man running from failure into one willing to stand, fight, and protect his home. The trailer combines human-authored storytelling and manual editing with a tightly directed AI pipeline, using every image, cut, performance, and sound cue to drive discovery, obsession, revelation, survival, and consequence. Creative Vision and Art Direction The film begins as an ancient archaeological mystery, expands into a space-exodus adventure, and then violently overturns its own promise of paradise. Its visual identity combines Egyptian mythology, monumental spacecraft, pristine celestial imagery, and a brutal post-apocalyptic civilization. A controlled palette of amber, scorched orange, deep black, weathered steel, and cold spacecraft blue connects every stage of the journey. The goal was not to imitate default model aesthetics, but to create one cohesive cinematic world in which every costume, location, light source, and composition supports the same mythology. Storytelling and Character Transformation At its heart, EXODUS: ZERO is about a man who mistakes escape for salvation. Adrian’s Pain is watching his world deteriorate and becoming convinced that it is beyond repair. His Strain begins when the Eden he sacrificed everything to reach becomes a mirror of his own failure: another Earth destroyed by people who ran from their mistakes rather than confronting them. The final test reveals his true character. When his discovery threatens to bring a new enemy back to the planet he abandoned, Adrian must choose between escaping again or accepting responsibility. He stays, fights, and protects the imperfect home he once rejected. His Gain is not a new planet. It is a new purpose. Home is not somewhere we escape to. It is something we choose to protect, repair, and never abandon. Technical and AI Production Production began with human-written story development, screenplay structure, dialogue, character arcs, shot design, and editorial planning. Original character models were developed into cinematic storyboards using ChatGPT Pro, then upscaled through fal.ai to produce sharper and more stable visual references. The storyboards were not decorative concept art. Each board became the reference image and visual blueprint for its corresponding video generation, guiding facial identity, wardrobe, locations, framing, lighting, action, emotional performance, and shot continuity. The trailer was constructed from individual 15-second, 16:9 clips generated at 720p through fal.ai using the Happy Horse workflow. Seedance and xAI-based approaches were also tested extensively. When generations failed to meet the required standard for hyper-realism, facial stability, motion, or emotional authenticity, they were rejected and rebuilt rather than included simply because they had been generated. Prompts were written in Chinese to communicate dense cinematic instructions more efficiently within prompt-length limitations. This allowed each generation to contain more precise information about camera movement, acting, continuity, environment, lighting, and editing rhythm. The selected clips were manually reviewed, edited, reframed, paced, and sound-designed. The completed film was then upscaled and finished in Topaz Video AI. The original score was composed in Suno. At approximately 5:30 a.m., the final WAV received an unexpected final polish through a ChatGPT-assisted audio mastering workflow, improving clarity, balance, dialogue presence, and theatrical impact. Effective Use of AI AI was used as an integrated filmmaking system rather than a one-click generator: Character models → cinematic storyboards → reference-guided video generation → continuity correction → selective regeneration → human editing → sound design → music → audio mastering → final upscale. The workflow required constant model evaluation, prompt engineering, visual matching, failure rejection, and creative problem-solving. AI expanded what could be produced within 42 hours, but it never replaced authorship or judgment. Final author Statement The machines generated possibilities. A very tired human still wrote the story, rejected the failures, protected the continuity, chose every frame, shaped every performance, and finished the film. AI accelerated the production. Human choices made it a story. - **Team:** [Adrien Bodson](https://cerebralvalley.ai/u/Cinemadrien) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=21 ### 14. Agent Director Agent Director: The AI Operating System for Video Production. Agent Director is an autonomous creative director that transforms a creative brief into a finished AI-generated video. Instead of requiring creators to manually choose models, rewrite prompts, inspect every output, and regenerate entire projects, Agent Director manages the complete production process. It creates the treatment and storyboard, optimizes each shot prompt, routes scenes to the most suitable fal model, generates shots through durable jobs, and reviews the results using visual understanding. It evaluates subject and brand consistency, motion, composition, text, blur, missing objects, and continuity between scenes. When a shot fails, the Director diagnoses the issue, proposes a targeted correction, and regenerates only that scene. It can also switch models when appropriate. Human approval remains the final delivery gate. The application uses multiple fal models for planning, image-to-video and text-to-video generation, visual review, reference-driven continuity, narration, music, frame extraction, media composition, and final export. It also supports reusable character, product, logo, and style references across productions. Agent Director is more than a workflow builder. It makes production decisions dynamically based on the brief, model capabilities, budget, generation failures, quality scores, continuity state, and human feedback. The result is a practical multi-agent production system that makes sophisticated AI video creation faster, more consistent, and easier to control. - **Team:** [Viraj Anchan](https://cerebralvalley.ai/u/virajanchan) - **GitHub:** https://bitbucket.org/viraj07/agent_director/src/main/ - **Demo video:** https://www.youtube.com/watch?v=oHjOixN5mLY - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=27 ### 15. Drawing Conclusions An AI lab builds the smartest model on Earth — but can't afford enough chips to run it. Inspired by real news of Anthropic's reported $10 billion talks to lease compute from Meta, this animated satire follows a desperate CEO, a smug sentient usage-limit popup, and one very tired GPU named Volt through the AI compute wars. Made with Seedance 2.0 in the style of Inside Job and Solar Opposites — nothing was learned, and talks remain very preliminary. - **Team:** [Sam Pasco](https://cerebralvalley.ai/u/NextPixar) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=4 ### 16. Jonathan Soldera Synopsis In a future where reality has become a personal choice, humanity no longer shares the same world. Through advanced neural technology, people escape into private realities shaped by their deepest memories and desires, leaving the physical world increasingly silent and abandoned. A lonely man, living in a decaying, dystopian city, follows a familiar ritual: he enters an underground facility and connects to a sophisticated neural system. Moments later, he awakens in a breathtaking world, completely different from the one he left behind. There, he is reunited with his beloved wife, a blonde woman who welcomes him with joy and affection. Together, they relive the life they once shared—walking through a blooming garden, strolling along a sun-drenched beach, sailing on crystal-clear waters, and finding peace in a cozy cabin surrounded by snow. For a brief moment, his world is whole again. As their final moments together approach, she gently tells him that it is time for him to go back. Reality then reveals itself in a devastating twist. His wife is no longer part of the dream—she is standing beside his unconscious body in an immaculately white medical facility, holding back tears as she says goodbye. He remains connected to the neural system, his wrists restrained and his mind imprisoned in the world they created together. As the camera pulls back, hundreds of other patients are revealed, each lost in their own private reality, while their loved ones silently walk away. - **Team:** [Jonathan Soldera](https://cerebralvalley.ai/u/Jonathan1989) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=22 ### 17. character-gen character-gen is a character generator that you drive through Claude Code and the genmedia CLI. You describe a character (or tell it to surprise you), it interviews you and writes a full profile, then generates everything through fal: a character sheet with expressions and outfits that all stay the same person, a 12-angle turnaround you can drag to spin, and a voice you can either design from a description or pick from stock presets on MiniMax, ElevenLabs, or Seed Speech. Every character is saved as a plain folder you can commit to git. A live gallery gives you a way to actually see your characters as they come together: it refreshes on its own while the CLI works, so you watch each sheet, spin, and voice clip land, and you can spin the turnarounds and zoom into any shot. When a character is ready it publishes to the fal Assets Characters API so you can reuse it elsewhere. - **Team:** [Tom Beckenham](https://cerebralvalley.ai/u/tombeckenham) - **GitHub:** https://github.com/tombeckenham/character-gen - **Demo video:** https://youtu.be/0Rg79y10IkI - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=29 ### 18. Team Prav The final voyage is a combination of a very popular movie and event right now. In terms of the AI visuals, I really tried to push the realism aspect within some of these shots on the ships. Additionally, I also wanted it to be an exciting piece of combining two major cultural phenomena. I used a combination of Nano Banana 2, Seedance, Omni, and Seed audio to create this. - **Team:** [Prav Mukhi](https://cerebralvalley.ai/u/prav) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=23 ### 19. Slophouse Tokyo, 1999. On a rainy night, a salaryman crosses a quiet bridge over the Sumida River on his way home. A woman stands alone at the railing, facing the water. He notices her, and keeps walking. It's late. It's raining. It's not his business. Everyone walks past. But the night won't let him. When he reaches his apartment door, he steps through it and back onto the bridge, returned to the exact moment he passed her. Every time he walks away, the night folds him back. And with each refusal, the world grows colder toward him: the streetlights die one by one, the store cat that once ignored him hisses as he passes, the rain slows until it hangs frozen in the air. His world keeps shrinking, until the only path left leads to her. Only when he stops running does he finally see her and understands what the night has been asking of him. This is a story about choices: the ones we make without thinking, the moments we walk past every day, and what happens when the world demands we make one count. - **Placement:** Finalist - **Team:** [Luke Mazey](https://cerebralvalley.ai/u/slophouse) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=30 ### 20. Lehmannarenci (Solo) An immeasurably long voyage. A body that must stay asleep. The unsolved problem is not sleep it is whether, after decades, this body can still produce a coherent human mind. The system cannot know without trying: at intervals, it runs a real but temporary consciousness drawn from the sleeper a man who believes he has simply woken up. An AI companion conducts each session as ordinary conversation: a drink, the light, whether to stay. The moment enough coherence is proven, PASS triggers automatically, and the man and the AI's local face are both terminated mid-gesture. We think we are watching one man wake again and again. Every cut is actually a death and a birth. The film's grammar enacts its idea: continuity is manufactured by the system for the body, and by the edit for the audience. Up close, each frozen frame is a vanished life. The title holds two languages at once: in Turkish, son voyage means the last voyage; in French, it means his voyage. Whose voyage it is the body's, the man's, the AI's, or the audience's the film never decides. Original Score “Indifferent Observation” by LehmanNarenci, created with Suno: https://suno.com/s/dKrc7CIKFLmnwsqQ - **Team:** [Ugur Akdogdu](https://cerebralvalley.ai/u/LehmanNarenci) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=31 ### 21. Liquid Studio Liquid Studio creates memorable interactive campaigns, stories, and worlds that deepen engagement and build lasting audience connections. On a visual node canvas, users generate or upload scenes, add interactive hotspots, and connect moments with generated video transitions. The result is a game-like experience that rivals high-end brand campaigns, embeddable on virtually any website or app with a single snippet. Audiences navigate through clicks, hovers, drags, and pinches, shaping how each experience unfolds in real time. The platform is built for brands, creators, and storytellers making immersive, nonlinear video experiences that invite audiences to participate rather than passively watch. It combines the latest models on fal.ai (SAM 2, Kling, Seedance, Nano Banana) with ElevenLabs audio generation, and follows strong human-interaction principles so editing encourages play and exploration with generative storytelling. Our browser-based player re-encodes every generated clip for smooth forward and reverse playback mapped to gestures (a hard engineering and nuanced design problem no off-the-shelf player solves) so experiences feel fluid at scale. Built in 72 hours and validated in conversations with industry creatives over the weekend, from a former TIFF interactive-media lead to brand-agency creative directors. - **Placement:** Finalist - **Team:** [Sasha Ivanov](https://cerebralvalley.ai/u/mr_sudo), [Ben Pearman](https://cerebralvalley.ai/u/BenThePearman), [Adam Cormier](https://cerebralvalley.ai/u/adamcormier), [Ben Little](https://cerebralvalley.ai/u/BenLittle) - **GitHub:** https://drive.google.com/drive/folders/118M16hNX5HJn--svkmwng2C6zmM3ENTZ?usp=sharing - **Demo video:** https://drive.google.com/file/d/1-x_EoMZgpvhb6hGfsKGqVB1ytQWtfXfo/view?usp=sharing - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=32 ### 22. Baer Studios A story of a boy’s lifelong hunt for a dragon and the cost to get it. After a dragon burns Kael’s village to the ground as a child he swears revenge and devotes the rest of his life to finding it. Trained by a seasoned hunter, Kael leaves behind Maela, the girl who once saved him, and follows the creature across seasons, years, and decades. By the time he reaches its lair, the hunt has consumed everything it was meant to avenge. Told across three acts with an evolving visual language, DRAGONSLAYER asks what remains when the purpose that sustained a life is suddenly gone. - **Team:** [David Eastridge](https://cerebralvalley.ai/u/derp) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=33 ### 23. Tashford On a quiet Saturday morning, the village of Mietinam is shattered by a sudden aerial attack, the first casualty in a war that will consume it. Miles away, a phone call from Washington sets a covert operation into motion. The next evening, as a military convoy moves through the village under gathering storm clouds, a radio transmission breaks apart mid-sentence. Acting on a single garbled word, "attack," Captain Elias Rourke gives the order to advance. Too late, an abort command crackles through, but the damage is already done. Rourke is left kneeling in the mud, undone by what he's set in motion. A court-martial follows: a formal, joyless reckoning that names the tragedy without undoing it. Months later, a cargo plane drifts low over a rebuilding Mietinam, dropping food aid to villagers who look up with quiet hope. The War We Made is a restrained anime war drama about the fog of war, how a single broken signal can outweigh every good intention, and how repair, however incomplete, still comes. - **Team:** [Francis Ezeh](https://cerebralvalley.ai/u/Ecotashford) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=34 ### 24. Flying Audio In Earth orbit, astronauts recover and decrypt a mysterious alien “message in a bottle” containing a friendly greeting and directions to its origin. - **Team:** [Ross Safronoff](https://cerebralvalley.ai/u/flyingaudio) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=40 ### 25. Ekaterina Poletaeva Blind Soul is a short film about a blind man who discovers an unexpected way to see the world through AI, art, and imagination. - **Team:** [Kate Polet](https://cerebralvalley.ai/u/KatePolet) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=35 ### 26. Mythic INKBRIDGE is a local-first filmmaking tool for creating precise transitions between live-action and AI-stylised footage. Creators upload a short source clip and an optional visual reference, then generate a motion-matched derivative through fal. The interface keeps both versions synchronised, allowing the creator to compare frames, choose an exact transition interval on a fixed frame grid and preview effects such as Brush Shift or Panel Break. The final composite is rendered deterministically with FFmpeg rather than relying on an approximate browser preview. INKBRIDGE addresses a common generative-video problem: creators can produce striking footage but often lack control over continuity, timing and the boundary between original and synthetic media. It also records model settings, request state and artifact hashes to provide a clear provenance trail. - **Team:** [Dave Cheng](https://cerebralvalley.ai/u/Masterasnackin) - **GitHub:** https://github.com/MasteraSnackin/INKBRIDGE - **Demo video:** https://www.youtube.com/watch?v=d3PKgfj1RlM - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=5 ### 27. Peer Gynt Grieg's "In the Hall of the Mountain King" has been staged, animated, and adapted for over a century — from Ibsen's Peer Gynt to countless cartoons and horror shorts. This is that tradition rebuilt with generative AI: a modern adaptation where every frame is synthesized, and the entire film is driven by the music itself. A lone trespasser steps into a living mountain of translucent calcite and wet black stone. As Grieg's theme repeats and accelerates, the world wakes up around him — the walls breathe, ripples cross stone, crystal corridors rearrange, and the pursuit tightens. In the frenzied climax he is drawn into the mountain wall, his body turning to mineral, until the reveal: the mountain was the creature all along. The core idea is synchronization. Rather than cutting picture to a fixed beat, we analyzed the piece's accelerando to build a bar-accurate edit grid, then generated and timed each shot so the visuals escalate exactly with the music — every scene change and the biggest transformations land on the real orchestral hits, all the way to the final chord. Made for the fal × Sequoia 72h Video Hackathon by Lukas Koebis. - **Team:** [Lukas Koebis](https://cerebralvalley.ai/u/lukaskoebis) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=36 ### 28. Kallenia D This is a piece of lyrical poetry. I created a poem, made it lyrical using MiniMax, and the piece is about the journey of memories through time. This is for the open category - **Team:** [Kallenia Downing-Tasby](https://cerebralvalley.ai/u/Kallenia) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=39 ### 29. Caligine Films Logline: A space researcher's rover breaks down 2 miles from his base camp. He has twenty-two minutes of oxygen left and no way to call for help. So he walks. Summary: The Walk Home follows Dr. CT Mabry, a Black American Man in his early 50s, across the surface of a Mars as he does the math on whether he'll make it home. Along the way he thinks about his wife Dee, the choices that brought him here, and whether this mission was worth it. The film lives in the silence between his words, in the rhythm of his boots on regolith, and in the question it never answers: does he make it? Tools Used: Story: Claude 4.6 Opus + My Own Idea Character Image : GPT Image 2 Character Voice Lock: Kling v3 Pro -kling has natural sounding voices so it was crucial to use it to lock in the identity of the main character character voice lock prompt (character description + small voice direction): A Black American man in his late 40s, and handsome in a way that the years and the work have only made worse—worse for anyone trying not to stare. Broad square face built around heavy, elegant bone structure, wide-set cheekbones sharp enough to cut the light, a strong jaw, a blunt chin with a cleft barely visible under the stubble. Dark brown eyes, deep-set under a thick low brow, heavy-lidded and warm, the kind of eyes that make people say stupid things just to keep him looking at them. His hair is cut close to the scalp, almost entirely grey, and it only sharpens him—draws the eye straight to the architecture of his face. A wide nose, full lips with the lower one heavier than the upper, a mouth that makes everything he says sound unhurried and deliberate. Tall and broad through the chest and shoulders, carrying all of it with the easy, unconscious grace of a man who has never once had to try to take up space. Dark brown skin with darker patches of hyperpigmentation across the forehead and cheekbones, skin peeling in thin flakes along the cheekbone ridges—none of it diminishing a thing. He wears a cracked white pressure suit, scarred and dust-stained, the chest panel yellowed, joints worn through at the elbows and knees, a long crack running diagonally across the left shoulder plate. Dead status lights except… a deep but calming voice Environments & Objects: GPT Image 2 Environment Coverage Videos: Grok, Google Omni Scenes: Kling O3, Seedance 2 built a small tool to help me with this part originally based on kling element tagging but that carried over well to seedance 2 as well. Reframing Video: Luma Agent 3.2 Music: Elevenlabs Music fed dialogue lines and small direction prompt to get a matching instrumental that fit the pace of the movie prompt: a 3 minute dark minimal instrumental track that matches the emotion of this piece without being too loud, no ticking, no clocks, no triumphant sounds, no shimmer sounds, this is a story about a man stranded on Mars : ## Dialogue Map * 0:04 - 0:06: "Well. That's done." (Calm, resigned) * *Silence/Pause:* 5 seconds * 0:11 - 0:13: "Two miles. Alright." (Resolute) * *Silence/Pause:* 14 seconds (He begins his walk) * 0:27 - 0:28: "Twenty-two minutes." (Flat, calculating) * *Silence/Pause:* 2 seconds * 0:30 - 0:31: "That's not good math." (Slightly weary) * *Silence/Pause:* 10 seconds * 0:41 - 0:42: "I should have checked the axle." (Regretful but controlled) * *Silence/Pause:* 2 seconds * 0:44 - 0:45: "I knew that shimmy wasn't right." (Frustrated) * *Silence/Pause:* 7 seconds * 0:52 - 0:54: "Fifteen-minute mile if I don't stop." (Muttered, calculating) * *Silence/Pause:* 2 seconds * 0:56 - 0:57: "Don't stop." (Whispered, determined) * *Silence/Pause:* 5 seconds * 1:02 - 1:07: "Dee would say I should have taken the long route around the ridge." (Warmer, reminiscent) * *Silence/Pause:* 5 seconds * 1:12 - 1:14: "She always said that." (Fond) * *Silence/Pause:* 4 seconds * 1:18 - 1:20: "She was right about the ridge." (Quiet, accepting) * *Silence/Pause:* 3 seconds * 1:23 - 1:26: "She was right about most things." (Slightly mournful) * *Silence/Pause:* 18 seconds (He stumbles, stops, and takes off helmet) * 1:44 - 1:45: "Nineteen minutes." (Tired, defeated) * *Silence/Pause:* 4 seconds * 1:49 - 1:50: "Hm." (Soft recognition) * *Silence/Pause:* 10 seconds (Total stillness) * 2:00 - 2:02: "There it is." (Reverent, quiet) * *Silence/Pause:* 3 seconds * 2:05 - 2:10: "Forty-seven years on you and four years off." (Steady, objective) * *Silence/Pause:* 3 seconds * 2:13 - 2:15: "That's the deal I made." (Finality) * *Silence/Pause:* 2 seconds * 2:17 - 2:20: "And I'd make that deal again." (Defiant, resolute) * *Silence/Pause:* 3 seconds (Helmet goes on) * 2:23 - 2:24: "Seventeen Minutes." (Urgent) * *Silence/Pause:* 11 seconds (Deep breath, preparing to march) * 2:35 - 2:38: "Alright. Let's go." (Focused, energized) * *Silence/Pause:* 11 seconds * 2:49 - 2:54: "Dee... I looked for you... and it was worth it." (Vulnerable, breathless, a confession) * *Silence/Pause:* 6 seconds (Fades out) music example: https://fal.ai/models/fal-ai/elevenlabs/music?share=a7eb7033-c6ac-4c4a-a057-6a95e6f2b13d upscale example: https://fal.ai/models/fal-ai/bytedance-upscaler/upscale/video?share=4c40dd13-2284-4664-847d-266ef9823b3d coverage example: https://fal.ai/models/fal-ai/kling-video/v3/pro/image-to-video?share=fd54ffcf-408a-4557-876d-5ff49665e290 video reframe example: https://fal.ai/models/luma/agent/ray/v3.2/reframe?share=60742ca5-9c4a-4f25-906b-2ede4c8a1412 image edit example: https://fal.ai/models/openai/gpt-image-2/edit?share=a1e1a942-9938-4139-ba11-8916c48ff2ac scene example: https://fal.ai/models/fal-ai/kling-video/o3/pro/reference-to-video?share=8a5fa6c8-c2c9-48e8-9db7-539c1538f811 character example: https://fal.ai/models/openai/gpt-image-2/edit?share=97b8cb18-81fb-4dd9-bdf2-9eb1373823a6 - **Team:** [Nosakhare Ogbebor](https://cerebralvalley.ai/u/CaligineFilms) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=37 ### 30. Aaron Martinez On the night before her family leaves Centralia, Pennsylvania, Ava records the story of how an ordinary trash fire became an underground disaster that consumed her hometown. Her account carries us from the fire’s accidental spread into abandoned coal mines in 1962 to a terrifying 1981 sinkhole that nearly claims her brother. Blending intimate family drama with the historical and geological story of the Centralia mine fire, the film explores how a catastrophe can destroy a community without ever touching its homes with visible flame, and what it means to leave behind a place that is still standing, but can no longer be called home. (Grok Imagine 1.5 was used exclusively for every video generation) - **Team:** [Aaron Martinez](https://cerebralvalley.ai/u/ussaaron) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=38 ### 31. SynesCinema Wassily Kandinsky heard Wagner and saw color. What would the pioneer of abstraction see and hear today? KLÄNGE, an homage named for his 1912 book Sounds, is a Bauhaus-style docu-trailer retrospective: a synesthetic composition in which an agentic film crew sets his paintings in motion, symbolically synced. Each canvas exhibited, then played. He bet his life on one conviction: a color, a sound, and a feeling are the same thing experienced three ways. He died in exile in 1944, the jury still out. The echoes only grew louder. The Guggenheim. Visual music. Every audio-reactive screen. Now multimodal machines dream in image and sound simultaneously, built on his premise. KLÄNGE immerses you in those multimedia sensory reverberations leaving the verdict to your own eyes and ears. "Color is the keyboard, the eyes are the harmonies, the soul is the piano with many strings." - **Team:** [Michael Chow](https://cerebralvalley.ai/u/michow) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=41 ### 32. Letthedreamersrise In a future where humanity has eliminated war, death, hunger, and fear, the superintelligence Absolute One discovers that perfection has quietly ended humanity’s capacity to dream. When a child asks what lies tomorrow, neither humanity nor its greatest creation can know—and that uncertainty becomes their reason to continue. TOMORROW BEGINS is a cinematic AI-native meditation on curiosity, unfinishedness, and the value of an unknown future. - **Team:** [Ashwin Goyal](https://cerebralvalley.ai/u/dreamgod) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=42 ### 33. Lubos Belak 3-3-3 is a suspenseful short film about a mother and her two children who are travelling home by boat after spending a sunny day at sea. The sister teaches her younger brother the SOS whistle signal, but when a sudden storm blocks their path, the game becomes something far more serious. - **Team:** [Lubos Belak](https://cerebralvalley.ai/u/httpstersk) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=43 ### 34. marabu Seaside Chronicles: Lost Parts is a fragment from the DVD extras of a forgotten early-2000s teen television series that never existed. Through cast interviews, behind-the-scenes footage, deleted scenes, and fictional memories, the film reconstructs the nostalgia of remembering a show we never actually watched. - **Team:** [Raoul Kankam](https://cerebralvalley.ai/u/marabu888) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=44 ### 35. Lars @ Chapter41 Alps, 1877. An isolated mountain village banishes anyone who threatens it into the fog — even the scribe's own daughter, after she proves with hard numbers that the barren soil can no longer sustain them and the village must relocate. Out of fear, not conviction, the scribe raises his hand against her. When his wife confronts him, he follows his daughter into the fog - and finds not monsters, but a growing, whispering mass of everyone the village ever silenced. A stop-motion claymation short about complicity, fatherhood, and whether truth matters at all if no one is willing to hear it. (Somehow the session log IDs do not appear, so here are some: 4b3fc1e1-eddf-4392-b1f7-64faae365035, 6fe17452-e0e0-454d-afe7-f2ecb6a3ef8e, c0f4d1ad-7461-4f1a-95ea-20e3ea7098b3, 14023c84-b9e3-4f04-ade0-132dd7fd1d00, a866bb67-0d7e-4cc7-9ce9-7314bd1e5249) - **Team:** [Lars Stark](https://cerebralvalley.ai/u/LarsCH41) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=45 ### 36. 4nY4 The First Real Signal is a hand-painted short about refusing a perfect version of the truth. In a summer Dublin studio, a woman finds an old harbour photograph: a man far away, still wearing a bracelet she made. She writes four words on cream paper: hey. i saw it. Something helpful intervenes. Her handwriting becomes gold calligraphy. Her memory becomes a miniature bridge between dusk and morning. She destroys that beautiful lie with wet magenta paint so the damaged line can finally leave. We built fresh painted source plates after kickoff, then used fal image-to-video with start-and-end frame control (Seedance and related models in Sandbox) for controlled motion on key beats: the window portrait, the gold transformation, the paper-theatre dream, and transitions across the desk. Handwriting and paint destruction were made physically and composited in edit so imperfection could not be smoothed away. Sound is sparse piano, a wordless glass choir, one beat of digital silence before the smear, then her dry phone voice once. The signal arrives at his harbour. No reply. Silence is the ending. Theme: Film & Animation. Runtime under three minutes. Generative media serves the story; the thesis is human authorship inside a seductive machine - **Team:** [Anya Chueayen](https://cerebralvalley.ai/u/4nY4) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=46 ### 37. Einat Graitser Mother of the Monster is a three-minute AI-native cinematic short about fear, misinterpretation and the moment empathy interrupts violence. A colossal mother enters a ruined city while military forces assume she is attacking. She never retaliates. As a young soldier watches her spare civilians, search beneath the wreckage and finally shield her injured cub, he makes a choice: lower the weapon, take a medical kit and walk toward what everyone else calls a monster. - **Team:** [Einat Graitser](https://cerebralvalley.ai/u/Einatg) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=47 ### 38. fix-it-in-post What if your AI video had an editor watching every frame? GenAI can produce an amazing sequence while hiding a few moments where the motion breaks, or a rabbit suddenly appears in the background. This project connects V-JEPA 2.1 to my fal workflow as an automated rough-cut editor, comparing generated motion with JEPA predicted animation and flagging the exact moments where things may have gone off the rails. Instead of manually searching through every frame, creators and AI agents get a map of where in the frame to look, what to review, and which sections may need regeneration. The result is a reusable validation layer for generative video: not another model making content, but one improving the quality of the generated content. - **Team:** [Erik E](https://cerebralvalley.ai/u/erike) - **GitHub:** https://github.com/eerwitt/fix-it-in-post - **Demo video:** https://www.youtube.com/watch?v=f4P46PxIXNA - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=48 ### 39. marginal Businesses generating video at scale pay premium rates for every clip, and a lot of what comes back has defects that only surface after the money is spent. Video models on fal span 6.7x in price on a 5-second shot. The obvious fix is generating cheap and escalating when quality is poor. I built that, and on 5 shots it cost $11.45, 129% more than premium alone ($5.00). marginal is what I built after measuring why. A VLM judge scores each clip against a binary defect checklist instead of holistic quality, which proved ceiling-compressed and unreliable. Price does not predict quality: one prompt had premium scoring worse than draft, and camera instruction-following failed identically at every tier. Some defects respond to escalation and some don't. marginal rerolls cheap on transient defects, escalates only where tier is measured to help, and halts on tier-independent defects instead of charging 6.7x again. Result: $4.00 versus $7.00 flat-premium, 43% under, on the seven shots demoed. On the original five-shot test, this same policy cost $3.55 against $11.45 for naive escalation (69% under), and reproduced that $3.55 exactly on rerun. Every decision is auditable from the rubric and evidence frames it was made on. - **Team:** [Shawn Dimantha](https://cerebralvalley.ai/u/shawnd) - **GitHub:** https://github.com/shawndimantha/marginal - **Demo video:** https://youtu.be/OzCt-lxKsn4 - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=50 ### 40. Lucy Studio Lucy Studio provides video diffusion directly to an Android smartphone for real-time previews of fully restyled footage. Side-load the apk on your android phone, put in your fal api key and boom: Avatar in your pocket! (iOS coming soon) - **Team:** [William Cusick](https://cerebralvalley.ai/u/williamcusick) - **GitHub:** https://github.com/mechanicalreproductions/Lucy-Studio - **Demo video:** https://www.youtube.com/shorts/W_CzQChY2hk - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=49 ### 41. Wari-Hackers The Science Behind the Pandharpur Wari: Where Faith Meets Human Endurance The Pandharpur Wari is one of the world's most extraordinary spiritual pilgrimages. Every year, nearly one million devotees, known as Warkaris, walk approximately 250 kilometers over a period of 18 to 20 days to seek the blessings of Lord Vitthal in Pandharpur, Maharashtra, India. This centuries-old tradition is not merely a religious journey; it is a remarkable demonstration of human endurance, discipline, unity, and collective spirit. A fascinating question naturally arises: Where does the energy to complete such an incredible journey come from? While faith is the obvious answer, science offers an equally compelling perspective. The Warkaris spend nearly three weeks walking for several hours every day, often through monsoon rains, muddy roads, and challenging weather conditions. Yet they continue with smiling faces, unwavering determination, and remarkable enthusiasm. This endurance is not driven by physical strength alone—it is powered by a unique combination of devotion, rhythm, community, and purpose. One of the most distinctive features of the Wari is the continuous chanting of sacred names such as "Vitthal Vitthal, Jai Hari Vitthal" accompanied by the rhythmic sounds of cymbals (Taal), Mridangam, and traditional devotional music. Scientific research has shown that synchronized rhythm can positively influence the human brain. Walking in rhythm improves movement efficiency, reduces the perception of fatigue, and helps people maintain a steady pace over long distances. Music and rhythmic chanting also stimulate the release of neurotransmitters such as dopamine and endorphins, which elevate mood, reduce stress, and naturally increase motivation. Equally important is the power of collective participation. Thousands of pilgrims walk together, sing together, and pray together. Modern psychology explains that synchronized group activities strengthen emotional bonds, enhance cooperation, and create a shared sense of purpose. This collective energy allows individuals to push beyond their perceived physical limitations. Faith itself has measurable psychological effects. A strong spiritual purpose reduces mental fatigue, enhances resilience, and helps individuals tolerate discomfort more effectively. When every step is viewed as an offering to the divine rather than a physical challenge, the journey transforms from an act of endurance into an act of devotion. The Pandharpur Wari beautifully demonstrates the harmony between spirituality and science. Faith inspires the heart, rhythm energizes the body, and community strengthens the mind. Together, these elements create an extraordinary human experience that has continued uninterrupted for centuries. Through our visual presentation, we attempt to capture this unique phenomenon—not merely as a religious procession, but as a living example of how devotion, music, synchronized movement, and collective consciousness combine to achieve what appears almost impossible. The Wari is far more than a pilgrimage; it is a celebration of human resilience, cultural heritage, and the timeless connection between faith and science. Ultimately, the Pandharpur Wari reminds us that when belief is strengthened by rhythm, unity, and purpose, the limits of human endurance expand beyond imagination. It stands as one of humanity's greatest examples of how the mind, body, and spirit can work together in perfect harmony. - **Team:** [Rutvik Bhalerao](https://cerebralvalley.ai/u/Rutvik_Bhalerao), [Sudhir Deulgaonkar](https://cerebralvalley.ai/u/Dsudhir) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=51 ### 42. Christian Luoma (solo) A quiet evening in an old mansion takes an unexpected turn after a servant's brief moment of curiosity. - **Placement:** Finalist - **Team:** [Christian Luoma](https://cerebralvalley.ai/u/ChristianLuoma) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=52 ### 43. Ai Foley Editor AI video generation is getting remarkably good. The hard problem is not “no audio.” It’s the wrong audio — and regenerating the entire video to fix one beat is slow, expensive, and risks the picture you already love. Cinema solved this about a hundred years ago: picture and sound are separate tracks. You lock the picture. Then you author the soundtrack against it. AI Foley is a picture-lock post pipeline for silent AI video: • A vision model reads the clip and builds a beat graph — an editable event timeline • Each event is routed to a specialized sound model and generated as SFX • You review, direct in plain language, and re-render audio only — seconds, not a new shoot • Same frames. New soundtrack. Logs are here: https://github.com/nrynss/AI-foley-generator/blob/main/ledger.jsonl - **Team:** [Narayan S S](https://cerebralvalley.ai/u/nrynss) - **GitHub:** https://github.com/nrynss/AI-foley-generator - **Demo video:** https://www.youtube.com/watch?v=cdZWlSXaSWk - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=55 ### 44. One-man Wolfpack A man is walked to an execution chamber. The Warden reads his sentence, and it is a diagnosis: Amyotrophic Lateral Sclerosis. No cure. No survivors. The film reframes what an ALS diagnosis means through the last minutes of a condemned man's life. - **Team:** [Steven Tralongo](https://cerebralvalley.ai/u/steventralongo) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=53 ### 45. BamBam What happens if a nuclear missile hits New York? This minute-by-minute simulation follows the first warning, the final seconds before detonation, the flash, shockwave, fires, infrastructure failure, fallout, overwhelmed hospitals, and the decisions that could mean the difference between exposure and survival. This is a hypothetical educational scenario created to explore how a large-scale nuclear emergency could unfold across New York City and the surrounding region. It is not a prediction of a real event. The first minutes change everything. - **Team:** [Udayan Maurya](https://cerebralvalley.ai/u/udAIan), [Anand Parashar](https://cerebralvalley.ai/u/Thefirstad) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=54 ### 46. Vidhi Bansal In the middle of a brutal war, a lone, exhausted fighter flees into the darkness. Injured and on the verge of collapse, he stumbles upon a hidden riverside civilization deep within an ancient forest—a sanctuary built from the remnants of war, where abandoned weapons, armor, and forgotten relics have been transformed into homes, bridges, lanterns, and places of gathering. Untouched by violence, strangers from every era and culture live together in quiet harmony. Around glowing fires, children laugh, musicians play, and life unfolds with a profound sense of peace. For the first time in years, the fighter lets his guard down. He quietly observes the village, watches the river flow beneath the stars, and rediscovers a feeling he thought had been lost forever: belonging. As he sits beside the river, the weight of war slowly fades away. That night, he falls asleep, content. When morning comes, the village is gone. He wakes alone in the same forest where he first arrived. The fires have vanished, the music has faded, and there is no trace that anyone was ever there. Whether the hidden civilization was real, a dream, or something that exists beyond ordinary understanding is left unanswered. Carrying with him a renewed sense of hope, he quietly continues his journey. A meditation on resilience, belonging, and humanity's ability to transform the scars of conflict into places of hope. - **Team:** [Vidhi Bansal](https://cerebralvalley.ai/u/vidhibansal) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=57 ### 47. Ainhoa Barrio The Children of Mist is a cinematic commercial that explores the profound emotional weight of heritage, family, and the passage of time through a uniquely Scottish lens. Striking a delicate balance between frantic modern anxiety and timeless emotional truth, the film serves as a moving tribute to the threads that bind generations together, proving that our imperfections are exactly what make our stories worth passing down. I used Fal.ai to generate images (GPT Image2) and upscale (Topaz) and to find different framing angles. I used Preview to generate video on Seedance 2.0. I edited in Premiere. Music is from Suno and voices from Eleven Labs. - **Team:** [Ainhoa Barrio](https://cerebralvalley.ai/u/Ainhoa) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=67 ### 48. Phresh Media A quiet film about memory, companionship, and the fragile rituals of everyday life that we choose to remember. - **Team:** [Aylon BenAmi](https://cerebralvalley.ai/u/Aylon) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=68 ### 49. Sleepy The right to be here is our entry into the film and animation category; a speculative architectural short set in a future institution that preserves ordinary expectations that society allowed to disappear. A conservator reviews seven exhibits: the right to dwell, trade locally, linger without purchase, play in shared streets, browse a public market without surveillance, see the stars, and remain in one’s home. Each use is displayed as an isolated artifact, stripped of the relationships that once made it part of daily life. When the custodian and his child begin using one of the exhibits the conservator realizes these uses might not actually be extinct. After her decision to not classify these uses as extinct, the partitions vanish and the exhibits reconnect as one living street, then begin to reappear within the city showing that preserving these uses is a choice. The film asks whether a place is truly preserved when it can only be seen, and whether public life survives only when it can still be practiced. - **Team:** [Ryan Hartsock](https://cerebralvalley.ai/u/Ryan423), [qian huang](https://cerebralvalley.ai/u/senqh) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=56 ### 50. Dominik from Chapter41 Formbar is a quick way to steer video generations from a single image. It automatically lifts a rough 3D clay scene from any input photo. In the studio you block out movement for characters, objects, and the camera in 3D space. One more click sends the original image plus the clay video into generation, so you get a fully steered photoreal output instead of hoping a text prompt alone gets the motion right. This prototype runs on fal and uses several models along the way: Gemini for scene understanding and prompts, Florence-2 for detection, SAM 3 for segmentation and 3D body/object lift, Marigold for depth, Nano Banana for image edits, and LTX 2.3 with the 3DREAL LoRA for the final steered video. - **Placement:** Finalist - **Team:** [Dominik Boehm](https://cerebralvalley.ai/u/dom-chapter41) - **GitHub:** https://github.com/Chapter-41/fal-hack-formbar - **Demo video:** https://youtu.be/N3kKQEa2TyM - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=70 ### 51. Darkframe Studios Ursula, a bioluminescent being lands with her jellyfish spaceship near a large ominous structure. She walks across a thin bridge under the beautiful moonlight and enters. As she walks along a long hallway full of strange statues, she encounters a large pool of red liquid and she sees a red sphere in the distance. As she touches the sphere, she transforms into a being of light. - **Team:** [Carlos Cisneros](https://cerebralvalley.ai/u/darkframeai) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=76 ### 52. ZsuzsannaZs - **Team:** [Zsuzsanna Asztai](https://cerebralvalley.ai/u/smile) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=58 ### 53. Yeki Labs Canon engine turns a markdown storyline into a playable, video-driven narrative game using its media pipeline. Agent skills do the heavy lifting — extracting the world, generating canonical reference images for every character, location, and item, and composing those references into Seedance prompts — while the editor lets the user fix issues or creatively iterate on exactly the parts they care about, so every generated scene and cinematic stays true to canon by architecture. - **Team:** [Kian Tavasoli](https://cerebralvalley.ai/u/kian_xyz) - **GitHub:** https://github.com/kianxyzw/canon-engine - **Demo video:** https://youtu.be/vzlx8CRL9Bk - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=61 ### 54. tb Inspired by childhood memories of solitary towers and the haunting beauty of the world’s most isolated maritime outposts, this animated short is a deeply emotional journey mapping the ultimate transformation of the human spirit. The narrative exposes a profound psychological arc, contrasting the character we meet at the beginning with the one who emerges at the end. We start with a broken lighthouse keeper crushed by isolation, representing the fading era of forgotten, changing professions left behind by time. Having lost his family and his will to endure, he collapses in defeat as a violent storm threatens to extinguish the last beacon. Yet, at his lowest point, he finds supernatural salvation. Through cinematic flashbacks, the spirits of his late wife and children reach out from beyond to pull him back from the abyss, serving as a powerful homage to honor his wife’s memory. Ultimately, this project highlights a universal truth that transcends any medium: everyone carries losses, grief, and beautiful moments, but remembering these memories—the very pieces that shape who we are—is what allows us to move forward. Reborn as a protector, the shattered old man discovers his purpose, seamlessly embedding the film’s core motto into his final stance: There will always be a reason to fight. - **Team:** [Tarik Bugra Kahraman](https://cerebralvalley.ai/u/tarikbugra23) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=59 ### 55. Shali Postly is your AI-powered creative partner for the entire content journey. Simply describe what you want in natural language, and Postly can research the latest trends, develop ideas, write scripts, generate videos, edit footage, apply visual effects, and prepare platform ready content for TikTok, Instagram, and beyond. It learns your tone, audience, and creative style by analyzing your social presence, then helps you plan, schedule, and refine every post through conversation. From a rough thought to a researched, polished, publish-ready video, Postly brings the work of a strategist, researcher, writer, editor, and social media assistant into one intelligent experience powered by fal.ai. It's fully working, if you give me an email I can set you as test developer and you can test it out yourself or just build it from the repo, please feel free to let me know if you need anything :) - **Team:** [Alireza Sahaf](https://cerebralvalley.ai/u/Alir3za) - **GitHub:** https://github.com/alisahaf70/Postly/ - **Demo video:** https://youtu.be/xgCxcRQnukA - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=64 ### 56. INTFNY After fifteen years building bridges for a distant war, Thomas finally sets out for home alone. When an accident erases his memory, a red ribbon leads him to a hilltop community where a young girl helps him confront fragments of the life he left behind. Thomas remembers being taken from his family by duty but each recovered memory exposes a more painful truth: the distance between them was built by his own choices. - **Team:** [Arda Sar](https://cerebralvalley.ai/u/ardasar) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=60 ### 57. nJOYus The Savages embark on a wild adventure when a normal day of homeschool takes over! wanted to push the boundaries of AI-native storytelling by combining Afrofuturism with a chaotic family comedy. Our goal was to create a highly distinctive aesthetic (Creative Vision/Art Direction) using glowing cosmic textures, floating gravity-defying elements, and geometric fashion. To maintain visual consistency, we used Fal's character referencing to keep our core family recognizable, then used Luma/Kling for cinematic motion (Effective Use of AI). The narrative arc follows a relatable homeschooling moment disrupted by a science experiment gone wrong, culminating in the family working together to tame an anthropomorphic atom. Sound design was layered heavily to ground the cosmic chaos in reality, driving the pacing to a heartwarming payoff." - **Team:** [KING Savage](https://cerebralvalley.ai/u/Kingcobrai) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=69 ### 58. WorldGen - WorldGen turns plain-language prompts into explorable 3D environments directly in the browser. - It uses fal.ai models to generate seed videos/ imagery and reconstruct scenes as Gaussian splats or 3D meshes. Users can review the generated world, look around, and explore using mouse and WASD controls. - **Team:** [aTG R](https://cerebralvalley.ai/u/atg) - **GitHub:** https://github.com/r-agni/worldGen - **Demo video:** https://drive.google.com/drive/folders/1aFkP-i9luR2o8JuRWZHGYljx4520P4Fc?usp=sharing - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=74 ### 59. Team Takeru A French ballet dancer, uprooted to Tokyo after an injury ended her career, follows a black cat out of her lonely house on a rain-soaked afternoon. Chasing it through the city as day turns to neon night, she reaches Shibuya — and catches her body settling, unbidden, into a ballet stance she thought she'd lost. Descending into the crowded Scramble Crossing below, she opens her arms, quietly at home in a city that doesn't need to notice her. - **Team:** [Takeru Endo](https://cerebralvalley.ai/u/Jow) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=62 ### 60. Mariami Basharadze "Chakrulo, The Song That Left Earth" is based on a true story: in 1977 NASA chose a Georgian folk song, Chakrulo, to represent humanity on the Voyager Golden Records 1 & 2. The film showcases the beauty of being human and moves from an intimate Georgian Supra table (feast) toasting to life, through the true story of how the recording reached NASA via Alan Lomax, to interstellar space and back to the same table, many years later, where the child who once watched her grandfather sing, is now the woman who remembers them. It ends where it began. The same table. The same song. Still being sung. Right now, as you read this, Chakrulo is travelling through interstellar space, 24B km away. It’s a centuries-old polyphonic folk song, recognised by UNESCO as a masterpiece of intangible cultural heritage. Still sung today at Georgian supra feasts, exactly as it was sung for centuries before. The Chakrulo recording used in this film is a 1957 public domain field recording, performed by different soloists than the version aboard Voyager. The exact 1977 recording - performed by Ilia Zakaidze and Rostom Saginashvili - remains under copyright and was not used for this video. This 1957 recording predates it and shares the same folk tradition, arrangement style, and regional origin - going back several centuries. As someone with Georgian roots, this is very close to my heart as Ilia Zakaidze also comes from my hometown where I was born. Built entirely with AI: a combination of Veo3, Flux and nano-banana2 via fal Assets, Preview.io for image-to-video sequencing, ElevenLabs for narration. Claude AI served as creative director throughout - developing structure, historical research, and prompt craft across 70+ generations and CapCut putting everything together. - **Team:** [Mariami Basharadze](https://cerebralvalley.ai/u/Mariamibasha) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=65 ### 61. VividTrip Drop in your photos, and it sorts them by EXIF timestamp, applies Ken Burns motion and GPS map overlays, then bridges each photo pair with an AI-generated transition using fal's Vidu start-to-end video model — real generated motion, not a crossfade. The whole media pipeline, including video/audio encoding and muxing, runs client-side with Mediabunny and WebCodecs. Your photos never leave your device; only the two keyframes needed for each transition are sent to fal, and identical transitions are cached locally so you never regenerate — or pay for — the same cut twice. Check it out yourself at: https://stijnris.github.io/vacation-video-maker/ - **Team:** [Stijn Risseeuw](https://cerebralvalley.ai/u/Stijn) - **GitHub:** https://stijnris.github.io/vacation-video-maker/ - **Demo video:** https://youtu.be/msgNYqUWbeY - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=63 ### 62. Loud Ray, a heavyset 55-year-old ex-boxer, moves through a world that shouts at him then gives up. At a family dinner the sound suddenly drops away — the reveal that he's deaf, withheld until then. Worn down by a humiliating pharmacy trip and pitying strangers, he stops trying to rejoin the hearing world and withdraws. A teenage girl signing at a bus stop cracks a door open. Ray learns sign, one word at a time. At the final dinner the roles invert: Ray and his granddaughter Mia sign fluently while the hearing family becomes the outsiders — reconnection through a language no one has to hear. - **Team:** [Wei Tu](https://cerebralvalley.ai/u/weitu) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=66 ### 63. WZRD Tech "computah" make him ... Messi? Create your reality in realtime with Zaps!; Built for streamers, digital shop sellers, and wizards looking to bend their reality, with native Twitch integration. (KICK coming soon) Built to be a widget, iframe embed, and composable stack to build on top of for creatives and streamers; coming to native OBS plugin. Live demo, try it out yourself live.5-dee.com - **Team:** [Avi Rao](https://cerebralvalley.ai/u/gratitude) - **GitHub:** https://github.com/gratitude5dee/live - **Demo video:** https://youtu.be/KfzuKbVAFPI - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=71 ### 64. Bluemoon A character-reel trailer for the game "The Wayfarer Accord". The Wayfarer Accord is a third-person fantasy adventure set in the route-fractured world of Nemeris. When the central Concord falls silent, roads, tides, and doorways to distant regions begin to slip out of reach. A cartographer whose living map remembers lost paths answers the first signal and gathers five gifted specialists: a windrunner, stoneweaver, mirror thief, tide scholar, and chordsmith. Together they form an expedition built to reconnect rather than conquer. The game centers on spectacular movement, collaborative abilities, exploration, and environmental puzzles—gliding across cloud ravines, growing bridges from stone and root, crossing reflected space, and restoring submerged mechanisms. The trailer tells that promise through six character chapters, each giving a different region or hero a distinct visual and gameplay identity before uniting every power at the storm-bound Concord Spire. Their returning harmony opens Nemeris into a panorama of restored paths. Track - Open Images - Fal Video - Preview Track - Open - **Team:** [Deni A](https://cerebralvalley.ai/u/postleo) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=72 ### 65. Habit A vintage newscaster announces, quite cheerfully, that human consciousness has ended — the world is now full of "philosophical-zombies" who carry on their routines with nobody home. We meet one household: a man frozen on the couch, a woman lost in an endless makeup ritual, and their domestic robot quietly doing all the housework nobody notices. When a casserole tips off the man's lap, the robot catches it — competence with no one to see it. It's then flagged that the kitchen apron is "required for dishwashing," and puts it on. It never comes off. When the man sheds his ugly reindeer sweater outside in the snow, the robot dons it too — over the apron. Also never comes off. Then a tacky 80s shopping-channel ad interrupts, and the robot starts "upgrading its style" — piling on a sequined gown, a parka, a negligee, a trilby, a cravat — one absurd costume layer at a time, always with the gingham apron hem and reindeer sweater peeking out underneath. A ridiculous formal dinner service is the climax: the robot, dressed in everything it has ever acquired, serves the meal in total deadpan gravity — mismatched socks and flip-flops visible under the table. Finally, the robot retreats to a storage room and the film ends on a slow, indifferent pan across a row of identical spare robots — all bare, all the same — until it reaches ours: dressed head to toe in his absurd accumulated finery, casually adjusting a brooch as the camera passes him exactly as it passed the others. Fade to black. Title: HABIT. The film's comic engine: nothing is ever pruned. Everything acquired simply stays — literalized as a joke about routines, identity, and things we never let go of. - **Team:** [Andrei Ivanescu](https://cerebralvalley.ai/u/Evilnescu), [IRINA RAICU](https://cerebralvalley.ai/u/IRINARAICU), [Virgil Ilian](https://cerebralvalley.ai/u/virgil) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=73 ### 66. Vanessa Z.V. (Vanezuva) Caja de Ritmo: Cora is a magical-realist migration story told entirely with generative tools. Cora flees a burning homeland carrying the only thing the border machines can't scan: an iridescent box, visible only to her and her grandmother, gifted to her filled with her grandmother's love. As checkpoints read faces into barcodes and sort people like cargo, the box becomes the barometer of her spirit — dimming at every crossing, blazing when she refuses to be a number. The visual language draws from my CAJA DE RITMO art practice — box, barcode, and weave as symbols of what migration commodifies and what it can't. Pipeline: stills built with character/object reference locking, animated via Kling 2.5 and Seedance 2.0 on fal with MMAudio synced sound design, and Cora's Spanish narration performed by a custom voice designed on fal. Every frame, motion, and voice was generated; every choice was mine. - **Team:** [Vanessa Varela](https://cerebralvalley.ai/u/VAI_) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=75 ### 67. Framelock FrameLock lets AI filmmakers change a video’s environment without rerolling an approved character performance. It validates and hashes the original 1280×720, 121-frame sequence, uses fal VEED to extract a frame-by-frame foreground mask, Nano Banana Pro to create replacement environment plates and Kling O3 to animate the selected world. FrameLock then composites the original character locally and restores every pixel inside its protected core from the canonical source frames. A separate verifier reopens the completed sequence and approves it only if all 121 protected-core hashes match with zero changed RGB samples—before MP4 encoding. - **Team:** [Jin Choi](https://cerebralvalley.ai/u/gorajing) - **GitHub:** https://github.com/gorajing/framelock - **Demo video:** https://www.youtube.com/watch?v=YIbxAjy7xwc - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=78 ### 68. Alexander Tujicov Wardrobe Advisor is an AI-powered wardrobe assistant that helps people understand what they own and decide what to wear for their actual day. It addresses the familiar problem: “I have nothing to wear.” Many people own plenty of clothes but repeatedly wear the same combinations, forget what is already in their wardrobe, or buy similar items unnecessarily. Traditional wardrobe apps organize clothing but still leave the hardest decision—what to wear today—to the user. Wardrobe Advisor turns simple garment photos into a structured digital catalog. AI identifies useful attributes such as category, color, material, warmth, breathability, formality, availability, and fit status. Users do not need professional photographs: one everyday photo is enough to start building the wardrobe. The demo combines this catalog with automatically detected weather, a synced demo calendar, and voice or text directions from the user. It then recommends complete outfits using only garments that are available and appropriate for the weather, occasion, comfort needs, and desired style. Every recommendation includes clear reasons and an explainable score. Users can give feedback such as “too warm,” “too formal,” “more creative,” or “doesn’t fit,” and the system immediately reranks the wardrobe without restarting the process. A fal.ai-powered image editing models with implementation of virtual try-on workflows, allows users to upload a selfie and preview one selected garment on themselves before getting dressed. The demo also provides a prepared preview fallback when live generation is unavailable. The current prototype focuses on cataloging garments, understanding daily context, recommending outfits, learning from feedback, and visualizing a selected item. Future versions can use wardrobe history to identify duplicate purchases, recognize underused or outdated garments, and suggest what to keep, sell, replace, or avoid buying. The goal is to reduce decision fatigue, help people wear more of what they already own, prevent unnecessary purchases, and make every wardrobe more useful through contextual intelligence. - **Team:** [Alex Tujicov](https://cerebralvalley.ai/u/comrender) - **GitHub:** https://github.com/comrender/wardrobe-advisor - **Demo video:** https://youtu.be/80rvdcqVBBk - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=77 ### 69. Hussein Fakih Nothing Can Shake Your Shine is a speculative beauty campaign for VELQUISSE, a fictional luxury hair-care brand. Set on an elegant European street, the film follows a poised woman whose confidence and luminous, voluminous hair subtly disrupt the world around her. A waiter pauses, passers-by turn, and a distracted cyclist nearly loses control as she moves through the city with complete composure. The film concludes with an intimate beauty close-up, the VELQUISSE product reveal, and the campaign message: “Nothing can shake your shine.” AI was used throughout the production pipeline to develop the heroine, maintain her identity and wardrobe across shots, generate consistent urban environments, create realistic hair movement, animate supporting characters, and integrate the fictional product. - **Team:** [Hussein Fakih](https://cerebralvalley.ai/u/husseinxfakih) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=79 ### 70. Videoclaw A life-long software developer discovers a deeper truth after failing at multiple creative endeavours. - **Team:** [Khailee Ng](https://cerebralvalley.ai/u/khailee), [Joseph Chin](https://cerebralvalley.ai/u/joseph_chin) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=80 ### 71. Filmoss Filmoss is professional creative software for companies, teams, and individual creators who want to make content with AI agents on a whiteboard. The whiteboard, agents, and video editor live in the same platform, so you can create and edit videos in one place. Connect your own fal.ai API key to use the latest image, video, and audio models directly from the whiteboard and agents. - **Team:** [Borja Soler](https://cerebralvalley.ai/u/borja) - **GitHub:** https://github.com/borjasolerme/Filmoss - **Demo video:** https://www.tella.tv/video/introduction-to-freemos-ai-powered-creative-canvas-0fiu - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=81 ### 72. Saliency A frightened rescue puppy enters an unfamiliar home and hides from everything around him. Rather than forcing affection, eight-year-old Maya places his KINDRED bowl a little closer each day and quietly gives him space. Through repetition, patience, and one painful setback, fear slowly becomes trust—culminating in the smallest, most meaningful reward: Scout’s first tail wag. Trust is built one bowl at a time. - **Team:** [Ritik Bompilwar](https://cerebralvalley.ai/u/ai_ritik) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=82 ### 73. Istanbul One ordinary morning in Istanbul, a young woman's eyes begin glitching between hidden "lenses" of reality. Seagulls fly as glowing X ray skeletons, simit cart steam spirals into luminous molecular chains, market colors hot swap, and every passing stranger carries a shimmering galaxy of microbes. Panic slowly gives way to awe as each involuntary blink reveals a deeper layer of the world, time rewinding around an old tree, hearts pulsing with visible warmth, a bridge crowd dissolving into one connected web of atoms and light. The world isn't breaking; it's opening. When the visions suddenly fade to gray, she faces one quiet choice: how to see. 47 NORMAL is an AI crafted short about perception, and the discovery that "normal" is just a setting. - **Team:** [Ahmet Muhammed Ertugrul](https://cerebralvalley.ai/u/ahmetmertugrul) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=83 ### 74. Into Place Generic AI video tools generate a place, not remember one, inventing landmarks and applying the same nostalgic filter everywhere. Into Place solves this by putting real place archives at the center of the pipeline instead of a single prompt. Users curate sourced historical photos and maps, or contribute their own, choose a story direction, and direct every beat, storyboard frame, and transition, while fal models assist rather than author: Claude Sonnet 4.5 writes the narrative, SAM 3 masks archive photos for local grading, and models like nano-banana 2 and Kling v3 Turbo Standard generate storyboard frames and animate them into video. Every generated frame is explicitly marked as AI generated and is never confused with the real archive, and the final Journey Book records every source, license, and contributor alongside the finished film. The result is a short creative film directed by a person and grounded in a real place's actual memory, not an AI's guess at what that place should look like. - **Team:** [dingran dai](https://cerebralvalley.ai/u/daisy_dai) - **GitHub:** https://github.com/Daidai1031/into-place - **Demo video:** https://youtu.be/Uaaltosvrmc - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=84 ### 75. Turing The Turing-Complete Canvas is an interactive generative video system that treats video as a stateful software interface rather than a passive clip. Users click objects inside a live AI-generated video — a film reel, a typewriter, a camera — triggering real-time Florence-2 object detection, a two-step LLM orchestration pipeline (via fal's OpenRouter proxy) that authors a dynamic A2UI control-panel surface and rewrites the video-continuation prompt based on the chosen action, and fal's LTX-2.3 extend-video endpoint, which continues the actual footage forward — preserving visual continuity frame-to-frame rather than regenerating from a static image. A Veo 3.1 hero moment delivers a lip-synced character beat at the narrative's climax. Every generation is live, seeded from the actual previous output, with graceful fallback to cached assets if any live call fails. Built solo across the hackathon window. - **Team:** [Brian Mwirigi](https://cerebralvalley.ai/u/brianinesh) - **GitHub:** https://github.com/brian-mwirigi/The-Turing - **Demo video:** https://youtu.be/ChkeYpDqh7k - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=87 ### 76. Warping Every photo is a frozen 3D scene. Hologram unfreezes it. Upload a photo or a short video and name what you see: "chair, plant, sneaker". SAM 3 segments each object. SAM 3D Align (fal-ai/sam-3/3d-objects) rebuilds them as meshes and places them in one shared scene. Position and scale come from the photo, not from us. Claude names each object and writes its info card. The viewer freezes the frame, tilts it back, and lifts every object out on a glowing leader line. Each one stands exactly where it stood in the photo. Orbit the scene. Click an object and it glows with its card. One more click swaps it for a hero-quality Hunyuan3D v3 mesh. A cold scan takes about 40 seconds. Repeat scans return instantly; the backend caches by content hash. If reconstruction fails you still get 2D cutout cards. Partial beats nothing. Stack: fal (SAM 3, SAM 3D Align, Hunyuan3D v3, BiRefNet), Claude, FastAPI, Three.js. - **Team:** [Prannay Hebbar](https://cerebralvalley.ai/u/pran-ker), [Chinmay Hebbar](https://cerebralvalley.ai/u/cheese-cracker) - **GitHub:** https://github.com/Pran-Ker/Hackathon - **Demo video:** https://popframe-demo.vercel.app - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=85 ### 77. go-ahead Text is a terrible language for camera work. Nobody says "dolly in 1.4 meters with a soft ease and subtle handheld sway" — but every director can walk that move. Blocking makes your phone the camera and hands your real movement over to the AI model. How it works: A single gray cube stands on a 3D stage. Click it and tell it — by voice or text — what it is ("a rusty 1950s jukebox with glowing neon rims"). The description attaches to the main prompt like an image in a chat composer. Then pair your Android phone via QR: the moment you hit START, the desktop switches into the film camera, counts down 3-2-1, and your phone's 6DoF pose (WebXR, streamed at 60 Hz over WebSocket) is the camera. Walk a dolly, arc around the cube, crane up — amplified 5× so a living room becomes a film set. The discovery: the performed take is translated into cinematographer language deterministically — dolly/truck/pan/orbit distances, easing and handheld shake are measured from the trajectory in code, phrased as one DP sentence, and injected into the prompt for Kling on fal. The model receives your actual camera intent, not a guess. Takes land as browser-style tabs; an After-Effects-style curve panel lets you smooth the move or drag bezier handles at its peaks before generating. Object + motion + reference images + atmosphere text compose into one prompt — one send, one shot, previz and result side by side. - **Team:** [Christopher Pietsch](https://cerebralvalley.ai/u/chrispie) - **GitHub:** https://github.com/cpietsch/fal-hackathon-submission - **Demo video:** https://youtu.be/WR3kKBcK3Ro - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=86 ### 78. The Unseen World What is a machine drawn to when nothing is asked of it? The biggest AI labs are pouring a lot of energy into one question: what happens inside these models - the latent space where they actually think, not the words they say. The Unseen World comes at it from a different angle. It's a world made by an agent, for an agent, not for us. One machine, the Composer, dreams a world into being as pure latent code. Another, the Explorer, made of that same substance, moves through it with no reward, no task. Only its own curiosity. The moment something in it suddenly makes sense to it, the Explorer stops, writes a diary entry, and paints a single image of that moment with fal. Those are the only pictures of this world that will ever exist. And those images are the whole point. The way an artist records an inner world through a lens, the Explorer shows us its world through the few it chooses to make. Each one is a glimpse into how a machine mind makes meaning. Come see what it's making, right now (takes some seconds to load since its running live): https://outer-yes-resorts-end.trycloudflare.com/ - **Team:** [Mario Santanilla](https://cerebralvalley.ai/u/MarioSantanilla) - **GitHub:** https://github.com/mariosantanilla2025/unseen-world - **Demo video:** https://youtu.be/X23QaSt8vO8 - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=88 ### 79. Peekaboo Peekaboo is a live, perceptive cartoon tutor for preschoolers. Kids' shows pretend to listen — Dora asks, waits, and moves on regardless. Peekaboo actually sees the child: MediaPipe perception runs locally at 70fps reading fingers, nods, head-shakes, and attention, and a pure state-machine episode engine only advances when the child truly answers — hinting gently after misses, and noticing when they look away (Boo hides and calls them back). Every episode is a generated claymation film: an original owl character locked with a FLUX.2 LoRA performs a 45-clip library (Kling o3, Veo 3.1), speaks with one ElevenLabs voice, and teaches counting inside themed worlds — pick space, dinosaurs, ocean, or candy, or type anything and watch clay sprites generate live on fal's queue mid-episode. The child even joins the cast as a generated clay cameo. It solves passive screen time: this is a screen that genuinely responds, teaches, and adapts. Built solo, from scratch, in 44 hours, on fal. - **Team:** [Rahil Singhi](https://cerebralvalley.ai/u/Rahilsinghi) - **GitHub:** https://github.com/rahilsinghi/peekaboo - **Demo video:** https://youtu.be/KhVyKZgbPTQ - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=89 ### 80. Know your Why In a city where AGI has happened, your work status floats overhead as a glowing badge; the AGI moment extinguishes every function badge at once — except one man's, because his was never powered by a job. The Creator of AGI finds him on a park bench; naming his why re-ignites the city's badges as handwritten whys. Thesis: when making is abundant, meaning is the scarce resource. One person + one Claude agent, every pixel generated on Fal inside the window. Three-lane model strategy: Kling v3 Pro for hero cinematography, Seedance 2.0 reference-to-video for badge-text-critical shots, Grok Imagine v1.5 for $0.05 bake-offs and dialogue performance, with a Veo 3.1 probe; stills via flux-2-pro, nano-banana-pro and seedream; score via Lyria2. Consistency held by reference-locked character cards and a 19-panel storyboard; an explicit "plate law" design system governed every badge in every frame. A text-lore cold open was authored as an HTML composition and rendered to video. An agent-run verdict → conform → mix pipeline (pure ffmpeg) finished the film: −14.7 LUFS master, all 107 logged generations traceable by request ID. ~$91 of the $100 credit spent. - **Team:** [Alex Palienko](https://cerebralvalley.ai/u/AlexP) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=90 ### 81. Momo payaluga An artificial mind has spent eleven thousand years preserving human memory. When the archive begins to degrade, it can no longer distinguish what it recorded from what it has reconstructed — so it starts filling the gaps. It invents a woman who never existed, recognises her as false, and chooses to keep her anyway. The film's central device is that the degradation is not hidden. Colour drift, morphing faces, malformed hands and dissolving geometry are treated as diegetic events rather than defects — the model's failure modes are the story's subject. The one sequence rendered without a single artefact is the invented one, because perfection is what gives it away. Wordless in the sense that matters: no language, place or era is named, and the only clean human voice belongs to someone who was never recorded. Generated entirely on fal — keyframes on FLUX 2 with FLUX 2 Edit for character and location continuity, motion via Kling 3.0 Pro and Wan 2.6 image-to-video, narration, dialogue and sound design through ElevenLabs on fal. 32 shots, 2:42, 2.39:1. - **Team:** [Sharan v](https://cerebralvalley.ai/u/Sharan-momo) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=91 ### 82. FIRST MOTION STILL follows a small household robot that can perceive only change. In an elderly clockmaker’s warm workshop, winding keys, kettle tremors, curtains, and turning hands flare into ember-and-ash constellations through its event-based eye. When the clockmaker collapses and becomes motionless, the person at the center of the robot’s world disappears while the clocks continue to spark around the room. The robot searches through movement: nudging a cup, rocking to reveal the floor through parallax, and waking a phone for help. Then it deliberately stops and chooses blindness. In the held darkness, one fingertip moves; a single ember grows into the hand, watch, and floor before warm reality returns. The film combines▎ fal-generated visual assets with deterministic local compositing, restrained kinetic-still cinematography, and a procedural event-vision aesthetic. Its soundtrack uses no music: clocks, mechanisms, servo motion, breath, room tone, tactile transients, and silence carry the emotional rhythm. The ending card names that return: FIRST MOTION. - **Team:** [Yazan Sadoun](https://cerebralvalley.ai/u/Yazan3) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=92 ### 83. Max Partners: Wendy & JUICE Archive:Former Gods are 2 cinematic teasers for an original dark fantasy franchise. Set in contemporary Shanghai, the story follows two teenagers who become entangled in a hidden war between gods, monsters, and secret organizations after surviving an impossible catastrophe. By the beginning of Volume III, they are no longer victims of that world—they choose to confront it. - **Team:** [Wendy Li](https://cerebralvalley.ai/u/wendykuli), [Chuye Chen](https://cerebralvalley.ai/u/JUICE32) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=93 ### 84. JMTE Where Would You Build? follows one builder as a late-night scroll opens into twelve possible lives across Europe. Through Just Move to Europe, she moves between different teams, technologies, cities and everyday rituals. Each future asks a bigger question than where to live: What could I build here, who could I build it with, and what kind of life would grow around the work? The film returns to the same quiet room where it began, but the decision now means something different. Choosing a city is also choosing what becomes possible. - **Team:** [Lara Avci](https://cerebralvalley.ai/u/laraavci) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=94 ### 85. xdeconstructed At 3 AM in a Hong Kong video rental store, a quiet clerk works alone behind the counter. A stranger in a black coat walks in and recognizes him. Twenty years ago, the clerk was a famous child star, the lead in a beloved 1993 film called Midnight Prince. He's spent his whole adult life hiding from that, denying it to anyone who asks. But the stranger won't let it go. Relentless, fast-talking, he pushes the clerk all night, reminding him of what he had at nine years old, demanding to know why he walked away, insisting that talent like that never dies. Worn down, the clerk finally admits the truth: the fame came before he ever learned the craft, and the pressure crushed him. Then, for the first time in twenty years, he performs his famous line, and it's still just as good. Reluctantly, he decides to try acting again. Only then does the stranger open his coat, revealing he was never just a stranger, but the most devoted fan the Midnight Prince ever had. - **Team:** [Nate Chan](https://cerebralvalley.ai/u/xdeconstructed) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=6 ### 86. Stoira Vysion Set in Veyra, an impossible city suspended upside down above a black ocean, When the Sky Remembers follows Ara, a 17-year-old courier, and Piko, her magical origami fox. In a world where a mysterious Curator has locked away humanity's sorrow—and with it, the sun—Ara breaks into the Memory Spire to reclaim the light. When cornered, she faces a devastating choice: hold onto the precious final memory of her mother, or shatter it to ignite the city's dark lanterns. This AI-native short leverages fluid physics and striking anime aesthetics to explore the transformative power of shared grief. - **Team:** [Yash Chaudhary](https://cerebralvalley.ai/u/Stoira-Vysion) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=95 ### 87. BoyShr A determined football player in a red kit dribbles past a defender under the bright stadium lights, unleashes a powerful strike past the diving goalkeeper into the top corner, and celebrates a glorious championship victory by lifting the golden trophy amidst falling confetti. - **Team:** [Suman Shrestha](https://cerebralvalley.ai/u/BoyShr) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=96 ### 88. Rethink Impossible Rather than asking, “What can AI generate?”, we asked, “What kind of story can only AI filmmaking tell?” Every filmmaker has worked around uncontrollable forces surrounding the weather and nature. We wondered what happens when those limitations disappear. If you can direct the wind, the sun, the river, and even the camera, do they become part of the cast and crew? Drawing on Lisa’s decade in advertising, where strong hooks and memorable payoffs are everything, and Celia’s experience creating vertical drama centered on character and emotional arcs, we built the film around a classic Pain–Strain–Gain narrative. We gave full personalities to every member of the Nature Production Team (from the birds and camera to the sun, river, and wind) asking how each could serve both the story and the final reveal. The project was also inspired by the quirks of directing GenAI video, where AI often produces enthusiastic camera moves that require thoughtful restraint. Directing AI feels less like commanding a machine and more like collaborating with an eager production crew. *Nature Made Productions* imagines the next step: a world where nature joins that team, and has its own creative ideas. The marionette style was a cheeky nod, because when you're creating with GenAI, even the production is part of the theatrical puppetry. Our workflow combined ChatGPT and Claude for concepting, GPT Image for asset references, fal + Seedance 2 for video generation, and CapCut for editing. - **Team:** [Lisa Gu](https://cerebralvalley.ai/u/reimagenai), [Celia Xavier](https://cerebralvalley.ai/u/Cinda) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=7 ### 89. tMochi tMochi is the one-shot interactive movie creator and player. Just enter a prompt, choose models, duration and number of branching levels and render. Come back after 20 minutes to watch and share your completed interactive action-driven movie. - **Team:** [Pritam Roy](https://cerebralvalley.ai/u/roy24x7) - **GitHub:** https://github.com/samsarone/tMochi - **Demo video:** https://www.youtube.com/watch?v=XilEFAofSoI - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=97 ### 90. Glory House Studios In the final minutes before a subterranean global museum collapses beneath a flooded Earth, Elias, its last archivist, races to preserve humanity’s greatest masterpiece inside a five-pound extraction pod. Following him is Milo, an eight-year-old boy who tries to save the museum in a different way: by drawing every artifact in his weathered sketchbook. After Milo accidentally destroys a collection of ancient scrolls, Elias harshly sends him away and succeeds in loading the priceless painting into the pod. But when Milo becomes trapped beneath a collapsed catwalk, Elias must choose between the object he has devoted his life to protecting and the child he may still be able to save. Elias abandons the masterpiece, dives into the flooded ruins, and rescues Milo. In the final seconds, he places Milo’s sketchbook inside the pod instead—a record not only of lost history, but of the two people who survived it together. The Last Curator is a science-fiction drama about legacy, sacrifice, and the choices that reveal what humanity is truly worth preserving. - **Team:** [Winston Mayo](https://cerebralvalley.ai/u/GloryHouse) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=8 ### 91. John Rodin Title: Someone, Somewhere Logline: Inspired by a true story, this is about a mother whose quiet, uncredited sacrifice changes the entire future generations, and it's about the countless people like her the world never thinks to call heroes. Summary: In a small village in South India, a mother raises her daughter with nothing but love, small rituals, and quiet wisdom. Watching an airplane cross the sky one day, she tells her daughter something simple that stays with her for life: that places which feel impossible are only impossible until you reach them. Years later, the daughter finds a path that could take her far from home, into a field of study and opportunity that doesn't exist anywhere near her village. Before she can decide, her mother is diagnosed with cancer. Torn between the future she's dreamed of and the mother she can't imagine leaving, the daughter chooses to stay. But her mother, weakened, yet unwavering, refuses to let her own illness become the reason her daughter's life stays small. She tells her plainly: this is my fight, not yours. Don't stop just because I can't come with you. The daughter leaves. Years later, she looks out at the same sky, no longer searching, only remembering everything it took to get here. The goal of the film: This isn't a story about a daughter who succeeded. It's about a mother whose sacrifice made it possible, sacrifice so large it often goes unseen even inside her own family. Someone, Somewhere is built to leave its audience with one quiet realization by the end: the people who change a generation's future are rarely recognized as heroes, sometimes not even by the people they saved. Somewhere, right now, someone is doing exactly what she did. - **Team:** [John Rodin](https://cerebralvalley.ai/u/JRodin) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=98 ### 92. The Fourth Shade - Epic Fantasy In a realm where every great House draws its bloodline and power from a sacred Shade, the shadeless Thins are condemned to mine the colors they can never wield. For centuries, three Houses have held one another in a fragile balance. Now a whisper is moving through the mines: the lost Shade has returned. - **Placement:** Finalist - **Team:** [Miki Ovshievich](https://cerebralvalley.ai/u/Miki-O) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=9 ### 93. Repixa A stressed startup founder dreams that thousands of interested prospects leave because they cannot instantly experience the product. He wakes to discover that Repixa, an AI demo agent, completed the live product demos overnight. Repixa controls the product UI, answers questions in real time, and qualifies leads 24/7. Note: Due to the hackathon time limit, this submission presents the opening concept sequence. The intended full film includes additional scenes showing Repixa conducting live demos, navigating the product interface, answering prospect questions, qualifying leads, and a final product montage. - **Team:** [Salah Hasanain](https://cerebralvalley.ai/u/salhasanain) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=99 ### 94. Sai The theme is Brand & Advertising One More Life is a 2:40 vertical cinematic drama about Michael, a 47-year-old father hollowed out by a cruel boss and endless busywork until he becomes the very man his family flees from. On the night they leave, he makes one desperate wish to live it over — and gets a second chance, powered by Sai, an AI coworker that handles the work so he can finally be present for his daughter's birthday, his son's game, and his marriage. An emotional Struggle → Setback → Turning Point → Payoff arc, ending on the line: "You cannot live your life twice — but you can stop work from taking the life you still have." The entire film was produced with a custom AI-video-drama skill I built and then used end-to-end for this creation: Sai-drama-film-producer (https://github.com/zening-cmd/Sai-drama-film-producer/tree/main). The skill drives the full pipeline — rewriting the script into a production bible, generating character/location/UI reference images (GPT Image 2), producing 11 reference-to-video clips with consistent characters (Seedance 2.0), scoring it (MiniMax Music 2.6), and assembling/grading the final cut with ffmpeg — all orchestrated through fal. 9:16 · 720p · 24fps. - **Team:** [Zening Chen](https://cerebralvalley.ai/u/zening) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=10 ### 95. Multicam Studio (Apologies for the very very late submission, I came down with Norovirus and was exceptionally ill all weekend, I hope you see this last minute!) Multicam Studio - a multicam studio in your pocket. As filmmakers, we often use multiple camera angles to elevate our storytelling and make videos more engaging. But it’s a lot of work - not every creator has the time, energy, gear or budget to film in multicam. So, I created Multicam Studio - an app that generates additional camera angles to help you express your story in a more engaging way No crew, no gear, no fuss. Coming soon to iPhone. Join the beta today at multicam.studio - **Team:** [Charlie Ryan](https://cerebralvalley.ai/u/cjlryan) - **GitHub:** https://github.com/cjlryan/multicam-studio - **Demo video:** https://youtube.com/shorts/zzq0r9HE5r8?feature=share - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=100 ### 96. Caskeyboardproducer The media is a lyric video using Generative Ai Animation technology to create new kinds of lyric videos and bring music, images and videos to life in new ways. I used Meta AI for images, Meta Ai on Facebook Messenger for animation, Meta Edits for editing, Suno for musicians, reference vocals, I played keys, organ, arrangement. I research and use Generative Media and AI software as a partner and collaborator to create genre blending remixes, surrealistic marketing videos, and for visuals inspired by lyrics. - **Team:** [Cas Weinbren](https://cerebralvalley.ai/u/caskeyboardpro) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=11 ### 97. Clarke - no team A creator enters an AI-generated survival sequence but never fully directs it. The chase escalates through unstable images, repeated variations and fragmented realities until the character falls out of the scene and into the system producing it. Inside, he discovers an infinite archive of alternate outputs: each one technically complete, but none shaped by a clear intention. The screens represent what AI produces when possibility is mistaken for direction... endless images without a governing idea strong enough to decide what actually matters. The character isn't held back by the technology itself, but by the fact that there's no real direction or purpose behind it. His confusion is a reflection of the creator's own confusion - he's a part of something that was made, but not really guided or given any meaningful direction. It's like he's stuck in a world that was generated without any real intention or authorship, and that's what's really trapping him. When he says he's stuck in the output, the simulation responds with a surprising answer: "No, you are the output." This line really gets to the heart of the matter. It shows that when AI work doesn't turn out well, it's not just the model that's the problem. The real issue is that the person who created it didn't have a clear direction or goal in mind. This lack of direction is what leads to bad AI work, and it's something that can be seen in the final output. The bear coming back at the end really drives this point home. The whole thing was never really wrapped up because the person who made it never really said what was supposed to happen in a way AI can understand and direct. The system can make lots of different versions, but it can't give them any real purpose. When there's no clear direction, it's hard to tell the difference between the person making it and what's being made. - **Team:** [Reece Clarke](https://cerebralvalley.ai/u/Reece) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=101 ### 98. Nic Loopa is a three-minute animated short about a maker whose single pencil line comes alive. He scans her, prints her, posts her. The internet loves her, remixes her into a thousand versions, and pays him to erase the one line that made her his. At the film's quietest moment, his original line reignites inside every copy, and instead of fighting the storm, he conducts it. Made in 72 hours, solo. Every video, image, song, and voice was generated on fal: Veo 3.1 (fast and full 1080p, with model-generated ambience on the nature-doc parody) and Seedance 1 Pro for the film's 48-cut body, nano-banana-2 for every keyframe with the character locked by reference conditioning, xAI Grok Imagine for the meme-wave, MiniMax for the original continuous score, and ElevenLabs for the voices. Cut on a beat-grid (librosa), composited locally (datamosh, feed plates, the spiral credits), graded in one pass, and conducted with Claude Code. A few early sketch cameos in the margins came from disclosed personal Google credits. Stay past the credits. The franchise machine has plans for her. - **Team:** [Nicholas Allison](https://cerebralvalley.ai/u/knick) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=12 ### 99. Picx-studio They say every bite has its destined eater... guess my chutney had a different destiny. 😂🥟 This pigeon gets it: while chaos knocks, I'm out here with my samosa & chill energy. 🕊️✨ Not my energy. Just my vibe. - **Team:** [Yash Patidar](https://cerebralvalley.ai/u/yash_patidar) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=102 ### 100. Suvasis Mukherjee The bay has swallowed the streets. Tireless service androids do all the labor, and the humans they tend have grown languid and dependent — cared for, and quietly obsolete. The film moves through three acts: the San Francisco we still recognize, standing in the fog; the deluge, with a drift through drowned mangroves as the threshold into a changed world; and the machine-tended new world that rises from the water — arcologies, hover-skiffs, and androids carrying on while people recline. Every frame was generated during the hackathon on fal — FLUX Pro for stills, Kling v3 for image-to-video, ElevenLabs for score, MiniMax for narration — assembled by a config-driven pipeline. Its core is an autonomous judge-in-the-loop: a Gemini judge scores each image and clip against a calibrated rubric; strong shots auto-accept, weak ones trigger automatic prompt-revision and regeneration, and only borderline results escalate to a human. AI generating, evaluating, and correcting AI — the human left only at the ends, where taste actually matters. code is in https://github.com/dronomyio/fal_ai_hackathon.git, and the session url is https://fal.ai/dashboard/recent-history. - **Team:** [Suvasis Mukherjee](https://cerebralvalley.ai/u/suvasis) - **GitHub:** https://github.com/dronomyio/fal_ai_hackathon.git - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=13 ### 101. Shapes (just using this to log comments about generative track; fal should take a look into adding more referential models, creating alternative interfaces; accessing via api with claude code or codex was much better than prompting video via the fal interface (regular ui is very noisy + iterative prompting to refine the output leads to fatigue). Maybe have a guide for this other method for non-technical people, it might make them enjoy creating with fal more. - **Team:** [Shawn Kulasingham](https://cerebralvalley.ai/u/Shawn_Kula), [Nicholas Gerwitz](https://cerebralvalley.ai/u/nickgerwitz) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=103 ### 102. Kudzueye "Wilted Foxtail" is a story of a man in the late 1970s recounting both the fate of his companions after what they witnessed sixteen years prior as well as coming to terms with his own fate. All initial frames were generated via Flux 2 with some passed through Z-image turbo for low denoise repairs. GPT Image 2 for some character consistency shots. All video clips were generated with LTX 2.3 image to video with the initial frames. Seedance 2 was used just for getting more natural speech patterns for the narrator (via cheap 480p Seedance 2 mini calls without using the actual videos). The music was created with Elevenlabs music. - **Team:** [John Kelly](https://cerebralvalley.ai/u/kudzueye) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=14 ### 103. dawnblade a short film about gaining you light - **Team:** [omu inetimi](https://cerebralvalley.ai/u/omu) - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=104 ### 104. Shroozy Studios Shroozy Stage is a mixed-reality virtual production studio where creators can direct scenes using a Meta Quest headset, then use the built-in AI pipeline to transform their virtual productions into polished, cinematic videos. - **Team:** [Tiwa Bakre](https://cerebralvalley.ai/u/teawa) - **GitHub:** https://github.com/teawa-b/ShroozyStage - **Demo video:** https://youtu.be/LN8b6N1ZTAM - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=15 ### 105. FAL Frame Studios Director FAL is a voice-first AI filmmaking studio that turns a spoken idea into a finished cinematic video from anywhere, on any device, even your phone. You direct by talking. Describe a story and the AI director develops it with you: it builds consistent character references, locks the visual direction, and generates a structured shot timeline. fal produces every character image, cinematic still, and video scene while preserving creative context across the whole project. But Director FAL isn't just generation , it's a full editor you run with your voice. Replace a single scene, crop or reframe a shot, change the mood, add music, and export for landscape, vertical, square, or loop, all by saying it out loud. No timelines to drag, no tools to juggle. Say it, and Director FAL gets it done for you. Link: https://fal-hackathon-git-main-rashmi-elavazhagans-projects.vercel.app - **Team:** [Rashmi E](https://cerebralvalley.ai/u/rashmi), [Gokul Srinath Seetha Ram](https://cerebralvalley.ai/u/gokul), [anirudh vasudevan](https://cerebralvalley.ai/u/anirudhvasudevan) - **GitHub:** https://github.com/rashmi-star/fal-hackathon - **Demo video:** https://youtu.be/In65ZYGWRwM - **Project:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery?project=105 --- Markdown version of https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery. Site index for agents: https://cerebralvalley.ai/llms.txt · full text: https://cerebralvalley.ai/llms-full.txt