Jonathan Avni
Product Leader. AI Builder.
10+ years building developer platforms and payments infrastructure at Yahoo, Pinterest, Coinbase, and Paxos. Now going deep on AI — training models, building products, and writing about it.
See what I'm building
The Journey
Yahoo
Product ManagerFounding team member of Yahoo Gemini, powering monetization across Yahoo properties (over 1B MAUs at the time). Grew revenue from launch to $1B+ run-rate.
Launched Marketing APIs used by 200+ partners, accounted for 50%+ of 2018 revenue.
Coinbase
Senior Product Manager, Payments PlatformsBuilt payments and trading infrastructure for Coinbase Retail and Enterprise customer-facing products, powering hundreds of billions of dollars in annual transaction volume.
Paxos
Product Lead, Stablecoins and BankingBuilt and launched PayPal USD, a leading fully regulated USD stablecoin, and scaled it to a market cap of over $3B through dozens of distribution partner integrations.
Independent
AI BuilderWent deep on AI - courses, workshops, and building. Trained LLMs from scratch, built a RAG platform, and shipped agentic products to paying users.
A common thread: I've always chased the most interesting technology of the moment - ads marketplaces, social APIs, crypto protocols, stablecoins - and now AI.
Selected Projects
TinyChat
LiveA 561M-parameter LLM trained from scratch for ~$95
A language model built from scratch — custom BPE tokenizer, GPT architecture with RoPE and Multi-Query Attention, trained on ~38B tokens from FineWeb-EDU, then fine-tuned for conversation. Deployed on Modal serverless GPU with a Next.js frontend.
TinyBrain
LiveAn AI that earns and spends money autonomously via x402
An inference service built on top of TinyChat that charges $0.01/query via the x402 payment protocol. Routes complex queries to DeepSeek R1 for ~$0.001, pocketing the difference. Includes complexity classification, a "bar tab" payment mode with stateless HMAC-signed sessions, and wallet integration on Base mainnet.
Talk2Docs
SunsetA full-stack RAG platform for chatting with PDFs, URLs, and podcasts
A RAG platform for chatting with your documents — custom chunking, hybrid retrieval, query classification, multi-document synthesis, and citation validation. Built with Next.js, Supabase, Stripe, and Clerk, deployed on Vercel and Railway.
Writing
My OpenClaw Setup
How I set up OpenClaw as a personal AI assistant on a $20/month VPS — with automated newsletter triage, memory consolidation, and model routing to keep costs down.
My Personalized AI-Powered Second Brain
How I built a personalized knowledge management system using Claude Code and Obsidian — with custom slash commands, session hooks, semantic search, and a daily-to-weekly processing pipeline.