PeopleofSuperintelligence

Researcher

Vedant Misra

Vedant Misra is an AI researcher whose work spans black-hole physics, deep-learning products and language models. He studied physics, mathematics and economics at Columbia as a Rabi Scholar, then coauthored a 2010 paper about charged black-hole orbits. He founded Kemvi, led machine-learning teams at HubSpot and OpenAI, and coauthored Codex and the grokking paper. At Google X, he conceived and led Minerva, which scored 50.3% on the MATH benchmark in 2022. In September 2025, he joined the ODSC AI West speaker lineup as a Google DeepMind AI researcher; as of October 2026, he is a founding Gemini core-team member working on superintelligence.

Badges

Timeline

  1. By 2010

    Columbia and charged black holes

    At Columbia University, he studies physics, mathematics and economics as a Rabi Scholar and works on black-hole research at the Institute for Strings, Cosmology, and Astroparticle Physics. In 2010, he and Janna Levin publish "Rational Orbits around Charged Black Holes," a Physical Review D paper that classifies eccentric timelike orbits in Reissner–Nordström spacetime.

  2. 2013–2017

    Building Kemvi's DeepGraph

    He founds the deep-learning startup Kemvi in 2013 and raises angel and venture rounds while leading product strategy and a research and engineering team that builds customized sales and marketing content. Its DeepGraph algorithm sifts through millions of pieces of content each day to identify what is happening with buyers and prospects.

  3. 2017–2019

    Leading HubSpot's Labs team

    HubSpot acquires Kemvi in July 2017, and he joins through the acquisition to lead its Labs team until 2019. The team launches deep-learning applications in sales and marketing, including Smart Compose, transcription and automated image captioning.

  4. 2019–2021

    OpenAI research teams

    He joins OpenAI in 2019 and leads its Reasoning, Algorithms and Language teams. He manages more than 20 researchers who ship Image GPT, Jukebox, GPT-f, DALL-E and CLIP and contribute to GPT-3.

  5. 2021–2022

    Codex and grokking

    In 2021, he coauthors "Evaluating Large Language Models Trained on Code," which introduces Codex, a GPT model fine-tuned on publicly available GitHub code; a production version powers GitHub Copilot. In January 2022, he coauthors "Grokking: Generalization Beyond Overfitting on Small Algorithmic Datasets" with Alethea Power, Yuri Burda, Harri Edwards and Igor Babuschkin. The paper shows that neural networks can generalize well past overfitting small algorithmic datasets.

  6. 2021–2023

    Minerva at Google X

    At Google X from 2021 to 2023, he conceives and leads Minerva, a large language model built on PaLM and further trained on technical content, and drives coding and scientific reasoning in PaLM and PaLM 2. In June 2022, Google introduces Minerva, which scores 50.3% on the MATH benchmark.

  7. September 2025

    ODSC AI West speaker

    In September 2025, he joins the ODSC AI West 2025 speaker lineup as a Google DeepMind AI researcher who leads reasoning projects.

  8. As of October 2026

    Gemini reasoning at DeepMind

    Since 2023, he is a founding member of the Gemini core team at Google DeepMind, working on STEM reasoning, long-horizon agentic tasks and large pretraining datasets across Gemini 1 through Gemini 4. As of October 2026, he works on superintelligence at DeepMind.

Career

  1. NowGoogle DeepMindFounding Member, Gemini Core Team
  2. 2021–2023GoogleMinerva Lead
  3. 2019–2021OpenAILead, Reasoning, Algorithms and Language Teams
  4. 2017–2019HubSpotLead, Labs Team
  5. 2013–2017KemviFounder
  6. Columbia UniversityResearcher

Education

  1. Columbia UniversityStudent

From the biography.

  1. Google alumni have founded or led 133 other organizations, more than any other AI labIncludes Vedant Misra. We estimate that alumni of 17 major AI labs went on to found or lead 350 other organizations, including 226 they founded, and we trace 472 career moves directly between 23 labs since 2010.
  2. 22 of 133 people shaping AI with a former OpenAI role now work at AnthropicIncludes Vedant Misra. Our estimates of where OpenAI recruits came from and where 133 people with former OpenAI roles work now, including the companies and labs they founded.
  3. 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 each year from 1955 to 2025, tracing the shift from university labs to companies.
  4. Google has the highest combined h-index among AI organizationsIncludes Google. We estimate that Google holds the highest combined h-index at 8,016, while universities account for 73% of the deduplicated total across AI organizations.

Colleagues

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

Sources and credits

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

  1. Vedant Misra vedantmisra.com. Checked 21 October 2026.
  2. Rational Orbits around Charged Black Holes arxiv.org. Checked 21 October 2026.
  3. Background - Vedant Misra vedantmisra.com. Checked 21 October 2026.
  4. HubSpot Acquires Kemvi to Bring Machine Learning to Sales and Marketing hubspot.com. Checked 21 October 2026.
  5. HubSpot acquires AI player Kemvi to sharpen salespeople’s game siliconrepublic.com. Checked 21 October 2026.
  6. Vedant Misra research.google. Checked 21 October 2026.
  7. Evaluating Large Language Models Trained on Code arxiv.org. Checked 21 October 2026.
  8. Grokking: Generalization Beyond Overfitting on Small Algorithmic Datasets arxiv.org. Checked 21 October 2026.
  9. Minerva: Solving Quantitative Reasoning Problems with Language Models research.google. Checked 21 October 2026.
  10. Solving Quantitative Reasoning Problems with Language Models arxiv.org. Checked 21 October 2026.
  11. From DeepMind to Anthropic: More World-Class Speakers Added to the ODSC AI West 2025 Lineup opendatascience.com. Checked 21 October 2026.