🧠 GraphRag for Universal Connectivity

Professional networking based on deep interests, not addictive reinforcement.

Connect with software developers, AI/ML engineers and researchers through intelligent matching based on your digital footprint. Leverage your data to create highly relevant, data-validated connections.

First launching in the Bay Area
Privacy-first
500+ engineers in beta

The networking problem every engineer knows.

You're building cutting-edge tech, but your network doesn't reflect your interests.

"You've read 12 arXiv papers but don't know how it gets deployed"
"You solve edge-case bugs alone when someone 5 miles away has the solution"
"You want feedback but don't want to ping Discord lurkers"
"You show up to meetups and still leave without meeting someone working on the exact thing you care about"
WHAT HAYVIN SOLVES:

Natural language networking.

MK

Mike Kim

ML Engineer at NVIDIA • 0.8 miles away

Working on sparse attention mechanisms for transformer efficiency. Recent paper on arXiv about 90% sparsity with minimal performance loss.

1

Type natural language query

Ask about specific technologies, research areas, or collaboration needs in plain English

2

Receive paper recommendations

Don't have time to read all the papers? Get graph-based recommendations based on your footprint

3

Get verified, relevant engineers

AI matches based on GitHub activity, papers, and actual technical work

4

Book coffee/climbing/walk

Meet domain experts in person with suggested conversation starters

Data sources we unify

GitHubXRedditOpenReviewarXivResumeCalendar
POWERED BY HYBRID RAG ARCHITECTURE

Technical but accessible.

Graph + Vector RAG processing for intelligent matching.

Data Ingestion

GitHub, OpenReview, arXiv, résumé

Graph Processing

Neo4j knowledge graphs

AI Matching

Graph-powered recommendations

Real Connection

In-person meeting facilitation

FEATURES

Built for Engineers.

Signal-Rich Profiles
GitHub commits, arXiv papers, OpenReview contributions - see what people actually build
Hyperlocal Discovery
"Within 1 mile working on RLHF" - precise location-based technical matching
Real-World Focused
Supplements meetups, doesn't replace them - nudges people to collaborate in person
AI Conversation Starters
Skip small talk with personalized discussion topics based on shared interests.
Calendar Integration
Automated scheduling - no more back-and-forth to find meeting times
Privacy-First
Your data stays yours - transparent about what we use and how
JOIN THE BETA

500+ engineers already in our beta waitlist.

Be among the first to experience intelligent networking for technical professionals.

500+
Engineers in beta
50+
Companies represented
Bay Area
Starting location

We'll send you early access when we launch in your area

Ready to unlock your technical network?

Join the future of professional networking for AI/ML engineers. Connect with peers who actually understand your work - the models you’re building, the technologies you’re using. Leverage Hayvin’s graph to unify your personal data and power relevant collaboration, in a privacy first environment.

Bay Area • 2025 • Privacy-first networking