Memory · Recognition · Connection
AI-powered face recognition for networking events. Remember every connection. Never forget a name.
Features
Your AI-Powered Networking Copilot
Point your camera. FaceLink handles the rest — recognition, memory, enrichment, and follow-up.
Real-Time Face Recognition
AWS Rekognition detects and matches faces in under 500ms. Overlay cards appear instantly on your camera feed.
Persistent Memory
Every interaction is stored with mem0. Next time you meet, FaceLink recalls names, topics, and context automatically.
LinkedIn Enrichment
Automatic profile lookup with Firecrawl scraping. See someone's title, company, and work history right on the overlay.
Voice-First Interaction
Speak naturally. Gemini-powered transcription + agent routing understands "This is Sarah from Datadog" and stores it.
Self-Learning Loops
Three feedback loops — face confidence, memory retrieval quality, and intent routing — improve continuously with Datadog tracing.
CRM Meeting Context
After identifying someone, FaceLink asks "How did you meet?" — building a personal CRM of every networking interaction.
How It Works
Three Seconds From Stranger to Connection
FaceLink runs entirely on your phone's camera. No wearables, no glasses, no setup.
Point Your Camera
Open FaceLink on your phone. The camera feed streams frames to the backend every 2 seconds.
Face Detected
AWS Rekognition detects faces and searches against your personal collection. Known faces match instantly.
Identity + Memory
Name, occupation, LinkedIn, and past conversation context surface in a clean overlay card.
Connect + Remember
Tap to open LinkedIn. Add meeting notes. FaceLink remembers everything for next time.
Built With
Production-Grade Stack
Enterprise APIs, zero-latency inference, persistent vector memory.