Unique Research · Insight · November 15, 2023 · Adapted from Sequoia Capital (Sonya Huang) · Original: sequoiacap.com/article/partnering-with-factory-autonomous-ai-for-all/
Background: Factory, founded by Matan Grinberg and Eno Reyes, aims to automate software development with AI. Its “Droids” handle code review, refactoring, documentation, and code generation. The system sits on a “Droid core” (engine + algorithms for engineering data); a Reflection Engine adds safety and customization for third-party AI models. The suite lifts developer productivity and simplifies the SDLC. Amid concerns about wrong/unsafe AI-generated code, Factory plans to train more models toward a comprehensive engineering AI system. It recently raised a $5M seed from notable investors.
How Sequoia met the founders
Like any VC partner, Sequoia’s author gets many cold emails. Matan’s March note stood out from the subject line alone — “chat with an ex Level-7 physicist” — signaling a shared niche in high-energy theoretical physics. A quick search showed that his modest “collaboration with Juan” meant co-authoring a paper with Juan Maldacena (Institute for Advanced Study) while Matan was still an undergrad. Reply came in under an hour.
Days later on a walk, Matan was working at the physics–AI intersection at Berkeley, in a PhD, considering a leap to startups without a locked idea — but clearly wanting to build. High EQ matched the unusual background; Sequoia encouraged the leap. Within a week he was iterating on autonomous coding — what became Factory — with a co-founder who also had Sequoia ties.
Eno Reyes
Eno grew up in Atlanta; first met Matan at Princeton. He had just taken a Hugging Face ML-engineer role and moved to San Francisco’s Mission — one street from where his grandparents settled after emigrating from Puebla, Mexico, and opened a restaurant. Shortly before reuniting with Matan at a hackathon to start Factory, Sequoia’s team met Eno at AI Ascent (March gathering of leading AI thinkers). One year out of college and months into the new job, he already impressed enough to attend with Hugging Face CEO Clem Delangue.
The vision
Matan and Eno want AI’s historical momentum to unlock software creativity — not only among engineers. Facing persistent tech-talent shortages, many Fortune 500 firms still run decades-old infrastructure for lack of people to modernize it. Handing time sinks — code review, docs, testing, project management, debugging — to fully autonomous “Droids” / agents would be game-changing for engineering teams. They also see that for every software developer there are ~100 people without deep technical backgrounds who still have great ideas software could realize.
A year earlier that next-gen AI vision looked over-ambitious. Timing and fit now look right: deep ML/AI/physics expertise plus lightning speed — idea → signed term sheet, incorporation, and building in a day. Six months later they were delivering value to early customers including two public companies and two decacorns, with more complex features on the roadmap and a strong team forming. Sequoia welcomes them to the community and looks ahead to supporting a world where coding ability no longer constrains collective creativity.
— END —