PeopleofSuperintelligence

Researcher

Khyathi Raghavi Chandu

Also known as Khyathi Chandu

Khyathi Raghavi Chandu is an AI research scientist at Mistral AI, where she works on the Voxtral audio models and the Magistral reasoning models. Trained in computer science at IIIT Hyderabad from 2011, she earned a PhD in 2021 at Carnegie Mellon University's Language Technologies Institute with a focus on grounded multimodal long-form generation. After two years as a research scientist at Meta AI, she joined the Mosaic team at the Allen Institute for AI in 2023 and co-authored the OLMo language model and the Dolma corpus, which both won best paper awards at ACL 2024. At Mistral, she co-authored Magistral, the company's first reasoning model, in 2025 and the Voxtral TTS speech model in 2026.

Badges

Timeline

  1. 2011–2016

    Dual degree at IIIT Hyderabad

    She joins IIIT Hyderabad in 2011 to study computer science and works with Prof. Dipti Misra Sharma in her first year; an S.N. Bose Scholarship takes her to the University of Pennsylvania for a summer as a research intern. By 2016, she completes a dual bachelor's and master's degree there, with a thesis on code-mixed question answering for Indian languages.

  2. 2016–2021

    A PhD at Carnegie Mellon

    From 2016 to 2021, she earns a PhD at Carnegie Mellon University's Language Technologies Institute, advised by Alan W. Black, Eric Nyberg and Yonatan Bisk, with a focus on grounded multimodal long-form generation and a thesis titled Effective Anchoring of Multiview Narrative Generation. In June 2021, she releases Grounding 'Grounding' in NLP with Bisk and Black, a paper comparing how NLP and cognitive science define grounding. Summer internships take her to Apple in 2019, Google AI in 2020 and Meta AI in 2021.

  3. 2019–2020

    BioASQ and Rising Stars in EECS

    In 2019, she is among the winners of the BioASQ challenge. UC Berkeley selects her as one of its Rising Stars in EECS in 2020.

  4. 2022–2023

    Research scientist at Meta AI

    She works as a research scientist in Conversational AI Research at Meta AI in California from 2022 to 2023. In 2022, her research aims at seamless communication between humans and machines across multiple modalities and languages.

  5. 2023

    Mosaic at the Allen Institute

    In 2023, she joins the Allen Institute for AI as a research scientist on the Mosaic team, in Yejin Choi's group. Her research areas there include making language and vision-language models more reliable by reducing hallucinations, and building AI agents that plan and carry out complex interactive tasks.

  6. 2024

    OLMo, Dolma and ACL awards

    In 2024, she co-authors OLMo, an open language model released with its training data and code, and Dolma, a three-trillion-token English corpus for pretraining research. Both papers receive best paper awards at ACL 2024. She also serves as a visiting professor at Jio Institute in 2024.

  7. 2025

    Magistral and Voxtral at Mistral

    At Mistral AI, she co-authors Magistral, the company's first reasoning model, trained with its own reinforcement learning pipeline and announced in June 2025. In July 2025, she co-authors Voxtral, Mistral's first audio model, released as Voxtral Mini and Voxtral Small under the Apache 2.0 license and able to handle audio files up to 40 minutes long.

  8. As of October 2026

    AI research scientist at Mistral

    She works as an AI research scientist at Mistral AI in Palo Alto, training models from 3B to 123B parameters for audio and reasoning. In March 2026, she co-authors Voxtral TTS, a multilingual text-to-speech model that generates natural speech from as little as 3 seconds of reference audio.

Career

  1. NowMistral AIAI research scientist
  2. 2024Jio InstituteVisiting professor
  3. 2023Allen Institute for AI (Ai2)Research scientist
  4. 2021–2023MetaIntern, Research scientist
  5. 2020GoogleIntern
  6. 2019AppleIntern
  7. University of PennsylvaniaResearch intern

Education

  1. 2021Carnegie Mellon UniversityPhD
  2. 2016International Institute of Information Technology, HyderabadBachelor's and master's degrees

From the biography.

  1. Of 943 people shaping AI who worked at Google, 446 work there nowIncludes Khyathi Raghavi Chandu. Our estimates of where the 943 people shaping AI who worked at Google and Google DeepMind work now: 446 at Google and 452 across 278 other organizations.
  2. Google alumni have founded or led 133 other organizations, more than any other AI labIncludes Khyathi Raghavi Chandu. 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.
  3. The Massachusetts Institute of Technology trained or employed 44 of the 187 AI major-prize winners, more than any other universityIncludes Khyathi Raghavi Chandu. The Massachusetts Institute of Technology trained or employed 44 of the 187 AI major-prize winners, more than any other university. We compare doctorate and 5+ year employment links, career timing and current workplaces for 187 major-prize and 1,111 best-paper winners.
  4. The median h-index is 48 at universities and 14 at frontier AI labsIncludes Mistral AI. The median h-index is 48 at universities and 14 at frontier AI labs. We compare publishing careers, recent citations, dual appointments and job moves across 2,420 university researchers and 308 frontier-lab researchers.

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 Khyathi Raghavi Chandu yet. Send us a free-licensed photo.

  1. Sister Act and the Banyan Tree blogs.iiit.ac.in. Checked 20 October 2026.
  2. CV | Khyathi Raghavi Chandu khyathiraghavi.github.io. Checked 20 October 2026.
  3. Dr. Khyathi Chandu - Jio Institute jioinstitute.edu.in. Checked 20 October 2026.
  4. [2106.02192] Grounding 'Grounding' in NLP arxiv.org. Checked 20 October 2026.
  5. Khyathi Raghavi Chandu khyathiraghavi.github.io. Checked 20 October 2026.
  6. [2402.00838] OLMo: Accelerating the Science of Language Models arxiv.org. Checked 20 October 2026.
  7. [2402.00159] Dolma: an Open Corpus of Three Trillion Tokens for Language Model Pretraining Research arxiv.org. Checked 20 October 2026.
  8. [2506.10910] Magistral arxiv.org. Checked 20 October 2026.
  9. [2507.13264] Voxtral arxiv.org. Checked 20 October 2026.
  10. [2603.25551] Voxtral TTS arxiv.org. Checked 20 October 2026.