Tataloka — redesign your room from a photo or video, with furniture you can actually buy. Live at https://www.tataloka.com THE PROBLEM Furnishing a room is a guessing game: Will that sofa fit? Which style works with what I already own? What will it all cost? Most AI room tools answer with a pretty image of furniture that does not exist — nothing in it can be measured, priced or bought. Tataloka closes that gap for Indonesian homes: every piece in a design is a real product from a real store, with its real size, photo, stock status and price, placed in a room sized to your space. HOW IT WORKS 1. Capture — upload a photo, several photos, or a 10–20 s video of the room (the browser picks the sharpest frames on-device; the video is never uploaded). Optionally enter the size measured with a tape. 2. Understand — SAM 3.1 finds the floor, walls, windows, doors and existing furniture. 3. Design — Muse Spark reads every view and plans a layout for the chosen style, room size and budget, picking each piece from a shortlist of real catalog products (strict JSON schema). It explains the design and says which existing furniture to keep. 4. Edit and preview — rearrange furniture in a 3D room or on a to-scale floor plan, swap or add products, restyle in one tap, give text or voice commands ("add a floor lamp"), and let Muse Image restyle your own photo for a before/after comparison. 5. Shop — a shopping list with store prices and product links, CSV export and a print-ready PDF design board. REAL DATA, NOT MOCK DATA - 236-product catalog: 228 IKEA Indonesia products (official name, photo, price, SKU, stock) plus verified marketplace listings for items IKEA does not sell. - Furniture is scaled to each product's real dimensions; room size is labelled AI estimate, typical or measured. - Live Tokopedia & Shopee alternatives via Meta web search — only URLs the search engine actually returned are shown, never model-written links or guessed prices. A PRODUCT, NOT A DEMO Accounts (email/password or Google sign-in), auto-saved designs and renders, monthly AI credits with automatic refunds on failed calls, an installable mobile web app (PWA), and English / Bahasa Indonesia. Built end to end by an independent developer: React 18 + Three.js frontend, Node.js/Express + SQLite backend, 131 tests against a fake Meta API, deployed in production on InstaCloud.
More from Meta Global AI Developer Hackathon
Cardo turns the world around you into Kurdish flashcards. Kurdish is spoken by tens of millions of people, but it has very few learning tools, and most…

oqed is an LLM-guided superoptimizer for x86-64. You point it at a C or C++ project and it does the rest: - builds the project and runs its workload…

Playtest Agent plays a game from screenshots, looks for bugs, then replays each report and has a judge check it. I make games alone. Solo developers and…

Open-source npm malware detection on Meta's models: a live monitor that flags new malware about a day before the public lists, a Muse agent that investigates…

Video2URDF turns one short phone video of a real articulated object (a laptop lid, a drawer, a desk lamp, a microwave door) into a measured URDF/MJCF a robot…

Glimmer Scope opens up Muse Glimmer 30B so you can see and steer the concepts it thinks with. I trained sparse autoencoders on nine layers of Glimmer's…
