# Nexus - The Living Campus Graph

- **Event:** [Meta Global AI Developer Hackathon](https://cerebralvalley.ai/e/meta-ai-developer-hackathon)
- **When:** Oct 26, 12:00 PM – Nov 4, 9:00 PM (EST)
- **Where:** Online
- **Team:** [Isha kookanapalli Math](https://cerebralvalley.ai/u/isha962), [Prakyat Prakash](https://cerebralvalley.ai/u/Prakx), [Anurag Lnu](https://cerebralvalley.ai/u/anurag2796), [shruti shivani](https://cerebralvalley.ai/u/ShrutiShivani)
- **GitHub:** https://github.com/nexus-research-ai/nexus
- **Demo video:** https://youtu.be/iejjJqIcyqI
- **Gallery:** https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery
- **Page:** https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/4966

We have LinkedIn to show where we’ve been, and GitHub to show what we’ve built—but higher education has no living system that shows what researchers and students are building right now. In universities like RIT and Berkeley, breakthrough ideas remain trapped in department silos and stale lab directories. Students and faculty only discover complementary work when papers are published 18 months too late, while researchers hesitate to share early-stage hypotheses for fear of idea theft.

Nexus transforms campus collaboration from an accidental encounter into an automated, privacy-first workflow. It turns static university publications, active pre-prints, and student portfolios into an interactive, real-time Knowledge Graph spanning 2,500+ research nodes and 20,000+ cross-disciplinary links.

Core Capabilities of Nexus:
1. Live Research Feed & Continuum: A campus-wide feed where researchers broadcast active projects, what equipment or skills they offer (e.g., 128-channel mobile EEG, cleanroom fabrication), and what datasets or collaborators they need.
2. Double Opt-In Mutual Match: Bilateral cryptographic matching connects researchers with complementary needs (e.g., federated learning expertise paired with mobile EEG hardware). Identities and proposals remain masked until both researchers explicitly accept.
3. Private Hypothesis Vault: Allows researchers to seal unreleased research concepts on-device with SHA-256 client-side Merkle proofs, securing timestamped Proof of Priority (#NX-8492) before public disclosure.
4. Proposal Studio & Multimodal Grounding: Integrates live split-view markdown drafting with automated claim-evidence auditing. Claims cite specific papers, figures, and table coordinates.
5. Federal Funding Radar: Ingests live NSF, NIH, and DARPA solicitations and aligns grant requirements directly with pod capabilities.
6. NexusAI Co-Pilot: An integrated reasoning assistant powered by Meta Muse that performs contextual paper retrieval, citation verification, and compliance checks without hallucinating evidence.

## More from Meta Global AI Developer Hackathon

- [424 Devs](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/4963)
- [Tolbert Innovation Hub](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/4964)
- [h=tvqo5](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/4965)
- [Jarvis](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/4967)
- [solehero](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/4968)
- [MiloBroker](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/4969)

---

Markdown version of https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/4966. Site index for agents: https://cerebralvalley.ai/llms.txt · full text: https://cerebralvalley.ai/llms-full.txt