~/wiki/entities

Founders and team

[ entity ][ updated 2026-09-21 ][ confidence high ][ jev-1.13.0 ]#team · founders · culture · hiring

TL;DR TypeSafe AI was founded by Diogo Almeida (CEO, ex-OpenAI, co-invented RLHF/InstructGPT), Sasha Sheng (COO, ex-Meta/FAIR), and Erik Gafni (CTO, repeat founder). The team works in person five days a week in San Francisco. Daniel Gafni is the Python SDK maintainer per pyproject.toml.

Facts

Field	Value
Founders	Diogo Almeida (CEO), Sasha Sheng (COO), Erik Gafni (CTO)
Team page	https://typesafe.ai/team — headline "Build Prod, Not God"
Location	San Francisco office near the Embarcadero station; in person five days a week
Backgrounds cited	OpenAI, Google Brain, Meta/FAIR, Stripe, Airbnb, Plaid, Docker, "and more"
Open roles	https://jobs.ashbyhq.com/typesafe-ai (team page "Open roles" / "Join us")
Python SDK maintainer	Daniel Gafni daniel@typesafe.ai (`typesafe-sdk` pyproject.toml)
Adapter maintainers	Erik Gafni erik@typesafe.ai, Daniel Gafni daniel@typesafe.ai (`system-one-adapter` pyproject.toml)
Press contact	Ke Deng, Chief of Staff, k@typesafe.ai (Business Wire release)

Founders

Diogo Almeida — CEO. Team page: "Diogo co-invented RLHF and InstructGPT, the methods that lead to ChatGPT and GPT4. Previously, he was at Google Brain. His research has helped shape modern language models, and he continues to push the frontier of AI systems." Fun fact listed: "🎮 Made it to the top 30 in League of Legends in NA and top 2 in Hearthstone." He is the author of Blog: Introducing System One Models & Jev (2026-09-15) and writes there: "At OpenAI, I helped build the methods that made language models useful at following instructions and talking with people. That work ended up as the research behind ChatGPT." He posts as @CompleteSkeptic and on 2026-09-20 shared personal working notes on building a Jev-centred coding agent (routing economics under the KV cache, per-query context selection, conditional AGENTS.md, security-aware routing); they are notes, not documentation, and are digested at community tier in Patterns: agent internals, context and coding agents.

Note a documented disagreement over this framing: the Substack reviewer Anthony Maio writes that Almeida "was an equal-contribution primary author of the InstructGPT paper and contributed to GPT-4," which he calls "more accurate than the company's 'co-inventor of ChatGPT' positioning in some of the hype material on X." The Business Wire release uses "former OpenAI researcher and co-inventor of RLHF/ChatGPT." See Press and third-party coverage.

Sasha Sheng — COO. Team page: "Sasha is an ex-research engineer from Meta/FAIR where she worked on the News Feed, AI Experiences, and AI Research. An avid builder, she has organized many hackathons and published work at NeurIPS and ECCV." Fun fact: "☕ Won 7 hackathons in 2023 and is addicted to caffeine and Barry's Bootcamp."

Erik Gafni — CTO. Team page: "Erik is a repeat founder (Ravel, multi-modal AI for dna-sequencing), an early employee at two unicorns (Invitae and Freenome), and an inventor with numerous publications and patents. He specializes in building production AI systems." Fun fact: "♟️ Placed 44th at US Chess Nationals in elementary school." He is listed as a maintainer of the system-one-adapter package.

Engineering contacts from the repos

These names appear in package metadata, not on the team page:

typesafe-sdk (Python): author "TypeSafe AI support@typesafe.ai", maintainer Daniel Gafni daniel@typesafe.ai.
system-one-adapter: maintainers Erik Gafni erik@typesafe.ai and Daniel Gafni daniel@typesafe.ai.
@typesafe-ai/sdk (npm): "author": "evinism" in package.json. No real name is given in the sources.

See typesafe-ai GitHub organisation and repos.

Culture and values

Quoted from the team page:

"We're a close-knit, flat team from OpenAI, Google Brain, Meta/FAIR, Stripe, Airbnb, Plaid, Docker, and more. We're backed by top-tier investors who share our vision for building the foundation of truly transformative AI."
"Our team works in-person five days a week in our San Francisco office near the Embarcadero station. We offer competitive salaries, equity, and benefits, as well as lunch and dinner."
"We love what we do and care about our work a lot. We strive for excellence and craftsmanship and won't stop until we get there."
"We want you to be able to do the most impactful work of your career at TypeSafe and help define our future as a company."

The five stated values:

Value	Stated meaning
Positive-Sum Games with Long-Term People	"When the team wins, we all win."
Thinking in Bets	"Good decision-making involves a lot of uncertainty. We prioritize long-term goals over short-term gains."
First Principles	"We want our decisions to always stem from clear reasoning, grounded in our values and culture."
Prioritize Learning	"a key to success is learning as fast as possible about ourselves, our team, our processes, our product, the market, and everything else involved with accomplishing our mission."
Passion	"We love what we do. We're looking for people who share that passion…"

Closing line of the culture section: "A working environment built for human beings… We build machines; we don't try to be machines."

TypeSafe AI (company) — the company
Manifesto: Composable AI — Build Prod, Not God — where "Build Prod, Not God" comes from
Blog: Introducing System One Models & Jev (2026-09-15) — Almeida's launch post
typesafe-ai GitHub organisation and repos — the repos these maintainers own

Sources