PeopleofSuperintelligence

Researcher

Olivier Teboul

Olivier Teboul is a French machine-learning researcher whose work covers differentiable sorting, learnable audio frontends and audio generation, and he is the Chief Technology Officer of the voice-AI company Gradium. A PhD candidate and Microsoft Research Scholar at École Centrale Paris under Nikos Paragios, he gave a talk on segmenting building facades at the 2010 IEEE Conference on Computer Vision and Pattern Recognition. Later he coauthored Fast Differentiable Sorting and Ranking in 2020 and LEAF in 2021, and in 2022 the OTT toolbox, AudioLM and the DEDAL protein-alignment model. At Google Research, he coauthored the 2022 paper that introduced DiffStride, the first downsampling layer with learnable strides, which won an ICLR Outstanding Paper award. He is a founder of Gradium, which launched in December 2025 with a $70 million seed round.

Now at
Gradium
Badges

Timeline

  1. By 2010

    PhD at École Centrale Paris

    By 2010, he is a French PhD candidate at École Centrale Paris and a Microsoft Research Scholar, supervised by Nikos Paragios in the Orasis team of the MAS laboratory. In 2010, he gives a talk on segmenting building facades with procedural shape priors at the 23rd IEEE Conference on Computer Vision and Pattern Recognition in San Francisco.

  2. 2020

    Fast differentiable sorting and ranking

    In 2020, he coauthors Fast Differentiable Sorting and Ranking with Mathieu Blondel, Quentin Berthet and Josip Djolonga, published in the proceedings of ICML 2020. The paper builds differentiable sorting and ranking operators as projections onto the permutahedron, the convex hull of permutations.

  3. 2021

    The LEAF audio frontend

    In 2021, he coauthors LEAF: A Learnable Frontend for Audio Classification with Neil Zeghidour, Félix de Chaumont Quitry and Marco Tagliasacchi. The frontend is fully learnable and outperforms mel-filterbanks on speech, music, audio events and animal sounds.

  4. January 2022

    The OTT-JAX toolbox

    In January 2022, he coauthors Optimal Transport Tools (OTT) with Marco Cuturi, Laetitia Meng-Papaxanthos, Yingtao Tian, Charlotte Bunne and Geoff Davis. OTT-JAX is a Python toolbox, built on JAX, that solves optimal transport problems between point clouds and histograms.

  5. April 2022

    An ICLR Outstanding Paper award

    At Google Research, he coauthors Learning Strides in Convolutional Neural Networks with Rachid Riad, David Grangier and Neil Zeghidour. The paper introduces DiffStride, the first downsampling layer with learnable strides, and is one of seven Outstanding Paper award winners at ICLR 2022, announced in April 2022.

  6. September 2022

    AudioLM for audio generation

    In September 2022, he coauthors AudioLM with ten others, including Neil Zeghidour and Marco Tagliasacchi. AudioLM maps audio to discrete tokens and casts audio generation as a language modeling task. Trained on raw audio waveforms, it generates coherent continuations of speech and piano music.

  7. December 2022

    DEDAL protein alignment in Nature Methods

    In December 2022, he coauthors DEDAL (deep embedding and differentiable alignment) in Nature Methods with Felipe Llinares-López, Quentin Berthet, Mathieu Blondel and Jean-Philippe Vert. The model learns to align protein sequences and detect homologs, and improves alignment correctness by up to two- or threefold over existing methods on remote homologs.

  8. Since 2025

    Co-founding Gradium as CTO

    In December 2025, the voice-AI company Gradium launches with a $70 million seed round led by FirstMark Capital and Eurazeo. Coming from Google Brain, he is one of its four founders and its Chief Technology Officer, at a Paris company that offers real-time, multilingual transcription and synthesis.

Career

  1. Since 2025GradiumChief Technology Officer, Founder
  2. 2022Google Research
  3. Microsoft ResearchMicrosoft Research Scholar

Education

  1. École Centrale ParisPhD

From the biography.

  1. 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.
  2. 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.
  3. 20% of people at companies who shape AI also hold a university role, led by Google and Stanford UniversityIncludes Google. We estimate that 539 of the 6,766 people with a known current organization (8%) hold both a university role and a company role, including 20% of people at companies and 13% of people at universities.
  4. Google alumni have founded or led 133 other organizations, more than any other AI labIncludes Google. 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.

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 Olivier Teboul yet. Send us a free-licensed photo.

  1. Olivier Teboul videolectures.net. Checked 22 October 2026.
  2. Segmentation of Building Facades Using Procedural Shape Priors videolectures.net. Checked 22 October 2026.
  3. [2002.08871] Fast Differentiable Sorting and Ranking arxiv.org. Checked 22 October 2026.
  4. [2101.08596] LEAF: A Learnable Frontend for Audio Classification arxiv.org. Checked 22 October 2026.
  5. [2201.12324] Optimal Transport Tools (OTT): A JAX Toolbox for all things Wasserstein arxiv.org. Checked 22 October 2026.
  6. ICLR 2022 Awards iclr.cc. Checked 22 October 2026.
  7. Congratulations to the 2022 ICLR outstanding paper award winners! - AIhub aihub.org. Checked 22 October 2026.
  8. [2202.01653] Learning strides in convolutional neural networks ar5iv.labs.arxiv.org. Checked 22 October 2026.
  9. [2209.03143] AudioLM: a Language Modeling Approach to Audio Generation arxiv.org. Checked 22 October 2026.
  10. Deep embedding and alignment of protein sequences | Nature Methods nature.com. Checked 22 October 2026.
  11. Gradium: Solving voice | Gradium gradium.ai. Checked 22 October 2026.
  12. Gradium - FirstMark firstmark.com. Checked 22 October 2026.