PeopleofSuperintelligence

Researcher

Karan Saxena

Karan Saxena is a research scientist at Netflix whose work includes large AI models from Meta and Mistral AI. In 2017, he worked on his undergraduate thesis on knowledge graphs under Partha Talukdar at the Indian Institute of Science, and in 2019 he earned a master's degree at Carnegie Mellon's Language Technologies Institute. After working as an applied scientist at Microsoft from 2020 to 2024, he became a research engineer at Meta in 2024 and co-authored The Llama 3 Herd of Models. He co-authored Magistral, Mistral AI's first reasoning model, in June 2025, and was also a co-author of the Voxtral audio chat models and the Devstral model for code agents in 2025.

Now at
Netflix
Badges

Timeline

  1. 2016

    A fellowship at IIT Madras

    In 2016, he is an Indian Academy of Sciences Summer Research Fellow, working with V. Kamakoti at the Reconfigurable and Intelligent Systems Engineering (RISE) lab at IIT Madras. He is also a Google Summer of Code student in 2016, working on the European MaRs Analog Station project for the Italian Mars Society under the Python Software Foundation.

  2. 2017

    Knowledge graphs at IISc

    As a senior computer science undergraduate in 2017, he works on his thesis on knowledge graphs under Partha Talukdar at the Machine and Language Learning (MALL) lab of the Indian Institute of Science in Bangalore. He gives a EuroPython 2017 talk in Rimini on ranking news summary bots by semantic document relatedness.

  3. 2019

    EWISE for word senses

    In 2019, he co-authors EWISE (Extended WSD Incorporating Sense Embeddings) with Sawan Kumar, Sharmistha Jat and Partha Talukdar for the annual meeting of the Association for Computational Linguistics. The model handles zero-shot word sense disambiguation by predicting over a continuous sense embedding space.

  4. 2019

    A master's at Carnegie Mellon

    He earns a Master of Science in Intelligent Information Systems at Carnegie Mellon's Language Technologies Institute in 2019. With Eric Nyberg, Teruko Mitamura and other co-authors, he writes Pentagon at MEDIQA 2019, a multi-task system for filtering and re-ranking medical answers that achieves the highest Spearman's Rho and Mean Reciprocal Rank in the MEDIQA question-answering shared task.

  5. 2020–2024

    Applied scientist at Microsoft

    He joins Microsoft as an applied scientist in February 2020 and becomes Applied Scientist 2 in August 2021, staying until April 2024.

  6. 2024

    Llama 3 at Meta

    In April 2024, he becomes a research engineer at Meta in Sunnyvale, California. He is among the co-authors of The Llama 3 Herd of Models, published in July 2024, which presents a family of language models for multilinguality, coding, reasoning and tool use whose largest version is a dense Transformer with 405B parameters.

  7. 2025

    Magistral, Voxtral and Devstral

    In June 2025, he co-authors Magistral, Mistral AI's first reasoning model, which the company trains with its own scalable reinforcement learning pipeline. He is also a co-author of Voxtral, two multimodal audio chat models presented in July 2025, and of the 2025 Devstral report on Devstral-Small, a lightweight open-source model for code agents.

  8. As of October 2026

    Research scientist at Netflix

    As of October 2026, he is a research scientist at Netflix, with research interests in machine learning, artificial intelligence, natural language processing and data science.

Career

  1. NowNetflixResearch Scientist
  2. From 2024MetaResearch Engineer
  3. 2020–2024MicrosoftApplied Scientist, Applied Scientist 2
  4. 2016IIT MadrasSummer Research Fellow

Education

  1. 2019Carnegie Mellon UniversityMaster of Science in Intelligent Information Systems

From the biography.

  1. Google alumni have founded or led 133 other organizations, more than any other AI labIncludes Karan Saxena. 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. 16% of AI career moves in the 2020s went from a university to a companyIncludes Karan Saxena. We estimate that 16% of AI career moves in the 2020s went from a university to a company; 29% of moves starting at a university went to a company, and 76% ended the university role.
  3. Among employers with 30+ full-time cases, OpenAI leads primary technical role H-1B pay at $310,000, and Amazon, Google and Microsoft file 57% of all casesIncludes Netflix. Across 440,669 certified US H-1B filings from 44 employers (FY2020–FY2026), OpenAI leads primary technical role base pay at $310,000 and xAI leads overall base pay at $290,000.
  4. Google had the most people shaping AI every year from 2014 to 2025Includes Microsoft. 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.

Colleagues

People named in this biography, and people with dated positions at the same organization in the same years.

Sources and credits

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

  1. Karan Saxena · EuroPython 2017 · Rimini, 9-16 July 2017 ep2017.europython.eu. Checked 20 October 2026.
  2. Karan Saxena - ACL Anthology aclanthology.org. Checked 20 October 2026.
  3. Karan Saxena - Language Technologies Institute - School of Computer Science - Carnegie Mellon University lti.cs.cmu.edu. Checked 20 October 2026.
  4. Karan Saxena | alphaXiv alphaxiv.org. Checked 20 October 2026.
  5. [2407.21783] The Llama 3 Herd of Models arxiv.org. Checked 20 October 2026.
  6. [2506.10910] Magistral arxiv.org. Checked 20 October 2026.
  7. [2507.13264] Voxtral arxiv.org. Checked 20 October 2026.
  8. [2509.25193] Devstral: Fine-tuning Language Models for Coding Agent Applications arxiv.org. Checked 20 October 2026.
  9. Karan Saxena - Google Scholar scholar.google.com. Checked 20 October 2026.