Inspiration
Networking and meetings often start with guesswork—who's in the room, what do they do, how should you start the conversation? Then afterwards, much of the context is forgotten. We wanted a tool that helps before, during, and after the conversation—giving you instant context, saving the discussion, and mapping out relationships.
What it does
Orbit transforms networking with real-time AI:
- Instant face recognition pulls public profiles the moment someone enters your view
- Persistent identity indexing remembers everyone across sessions with full conversation history
- Live transcription captures every word while generating smart follow-up suggestions
- Dynamic relationship mapping visualizes connections between people and shared interests
Every interaction becomes structured, searchable, actionable intelligence.
How we built it
This is technically insane for a 36-hour hackathon:
- Custom computer vision pipeline with OpenCV + DeepFace + FaceCheckID handling multiple faces simultaneously through parallel processing
- Real-time multithreading architecture achieving sub-second latency via WebSocket connections
- Persistent face indexing creates searchable memory—instant recognition with full conversation history when someone reappears
- Live conversation intelligence using Whisper + LLMs for real-time context analysis and follow-up generation
- Advanced relationship mapping with Sigma.js algorithms connecting people through shared interests
- Polished React interface hiding massive technical complexity behind intuitive design
Challenges we ran into
- Real-time performance while orchestrating multiple heavy AI models simultaneously
- Reliable face indexing across lighting conditions and multiple participants
- Complex data pipeline from video → recognition → scraping → analysis without latency bottlenecks
Accomplishments that we're proud of
- Revolutionary networking experience never built before—real-time social intelligence through computer vision
- Technical breakthrough: Parallel face recognition, conversation indexing, and relationship mapping working seamlessly
- Exceptional UX: Complex AI systems hidden behind interface anyone can use instantly
What we learned
- Real-time AI orchestration requires careful architecture—every millisecond matters for instant user experience
- Computer vision complexity multiplies in real-world conditions with lighting, angles, and multiple faces
What's next for Orbit
- Enhanced recognition accuracy under challenging conditions
- Privacy-first architecture with comprehensive opt-in controls
- Enterprise deployment for conferences and professional networking at scale
Log in or sign up for Devpost to join the conversation.