PeopleofSuperintelligence

Lab leader

Diogo Almeida

Also known as Diogo Moitinho de Almeida

Diogo Almeida is an AI researcher and lab leader whose OpenAI work helped shape instruction-following models and the research behind ChatGPT. He earned a B.S. in computer science and mathematics at Rensselaer Polytechnic Institute, developed medical-imaging algorithms at Enlitic, and researched at Google Brain before joining OpenAI. His papers include a learned-optimizer study, InstructGPT, and the GPT-4 Technical Report. As of October 2026, he is CEO and co-founder of TypeSafe AI, where he leads work on Jev, its first public System One model.

Now at
TypeSafe AI
Works
12
Badges

Publication numbers from OpenAlex, as of 19 October 2026. Citations and h-index are hidden because this OpenAlex record may be incomplete or mix in other authors.

Timeline

  1. 2013

    A Rensselaer computer science degree

    He earns a 2013 B.S. in computer science and mathematics at Rensselaer Polytechnic Institute.

  2. 2014–2017

    Enlitic medical imaging

    He joins Enlitic as a senior data scientist in 2014 and develops deep-learning algorithms for medical diagnosis.

  3. 2017–2018

    Research at Google Brain

    He works in research at Google Brain in 2017–2018.

  4. 2021

    A learned optimizer

    In 2021, he coauthors A Generalizable Approach to Learning Optimizers, which learns to update optimizer hyperparameters and generalizes across modalities absent from training.

  5. March 2022

    InstructGPT and ChatGPT research

    In March 2022, he coauthors InstructGPT, which fine-tunes GPT-3 with human demonstrations, response rankings, and reinforcement learning from human feedback. Human evaluators prefer its 1.3-billion-parameter model to the 175-billion-parameter GPT-3, and he helps develop the research behind ChatGPT.

  6. 2023

    The GPT-4 Technical Report

    In 2023, he coauthors the GPT-4 Technical Report, which describes a multimodal model that accepts image and text inputs and produces text.

  7. 2024

    From OpenAI to TypeSafe

    He joins OpenAI in 2020 and leaves in 2024; in August 2024, he starts TypeSafe AI to build models designed for decisions inside software.

  8. As of October 2026

    CEO at TypeSafe AI

    As of October 2026, he is CEO and co-founder of TypeSafe AI. In September 2026, he introduces Jev, TypeSafe's first public System One model, for structured decisions in software.

Career

  1. Since 2024TypeSafe AICEO and co-founder
  2. 2020–2024OpenAI
  3. 2017–2018Google Brain
  4. 2014–2017EnliticSenior Data Scientist

Education

  1. 2013Rensselaer Polytechnic InstituteB.S.

From the biography.

  1. Google alumni founded or led 87 other organizations, the most among the labs we compareIncludes Diogo Almeida. We trace 212 organizations founded or led by AI lab alumni, and 223 career moves between the labs.
  2. 13 of 79 people shaping AI with a former OpenAI role now work at AnthropicIncludes Diogo Almeida. Our estimates of where OpenAI recruits came from and where people with former OpenAI roles work now.
  3. Google has the highest combined h-index among AI organizationsIncludes Google. Our estimates of combined and median h-index at AI organizations, using current affiliations and OpenAlex records.
  4. Google had the most people shaping AI every year from 2014 to 2025Includes Google. Our estimates of which organizations employed the most people shaping AI from 1961 to 2025, with a year-by-year chart.

Colleagues

People with dated positions at the same organization in the same years.

Sources and credits

Photo: We have no published photo of Diogo Almeida yet. Send us a free-licensed photo.

  1. Rensselaer Student Startup Reaches Semifinals at 2015 Rice Business Plan Competition everydaymatters.rpi.edu. Checked 19 October 2026.
  2. Deep Learning in Healthcare Part 2: Future & Predictions blog.re-work.co. Checked 19 October 2026.
  3. Diogo Almeida: Bio, Work & Ideas ai.engineer. Checked 19 October 2026.
  4. A Generalizable Approach to Learning Optimizers arxiv.org. Checked 19 October 2026.
  5. Training language models to follow instructions with human feedback arxiv.org. Checked 19 October 2026.
  6. Team - TypeSafe AI typesafe.ai. Checked 19 October 2026.
  7. Introducing System One Models & Jev typesafe.ai. Checked 19 October 2026.
  8. GPT-4 Technical Report arxiv.org. Checked 19 October 2026.