·
Builders & Technical Leaders
Omar Sanseviero
Staff AI DevRel Engineer & Developer Experience Lead |Google DeepMind
Leads developer experience at Google DeepMind for Gemini, Gemma (180M+ downloads), and AI Studio. Previously Chief Llama Officer at Hugging Face where he launched Spaces (400k+ demos) and grew the Hub from 3k to 3M repos. Co-authored O'Reilly's 'Hands-On Generative AI with Transformers and Diffusion Models'. Co-founded AI Learners, a Spanish-speaking AI education community.
Biography
Deep Research
Omar Sanseviero is a Peruvian-born ML engineer and developer experience leader who currently serves as Staff AI Developer Relations Engineer and Developer Experience Lead at Google DeepMind, where he leads a global team empowering developers to build with Gemini, Gemma, and AI Studio. Previously he was Chief Llama Officer and Head of Platform and Community at Hugging Face (2021-2024), where he grew the Hub from 3,000 to over 3 million repositories, launched Hugging Face Spaces (scaling from 500 to 400k+ community-built demos), and built the Developer Advocacy Engineering and Moonshot organizations (25 people). Before Hugging Face, Omar worked as a Software Engineer at Google on Google Assistant and as a 20% Product Manager on TensorFlow Graphics. He holds a BS in Computer Science from Tecnologico de Monterrey (2014-2018) and an MS in Computer Science from Georgia Institute of Technology (2019-2021). Omar co-authored the O'Reilly book 'Hands-On Generative AI with Transformers and Diffusion Models' (published December 2024), co-founded AI Learners (a Spanish-speaking AI education community), and writes The Hacker Llama Substack covering open ML news. At Google I/O Connect 2025 he announced Gemma had surpassed 180 million downloads and 80,000+ developer-built variants.
open-source machine learningdeveloper experience for AI productsgenerative AI with transformers and diffusion modelsHugging Face ecosystem and SpacesGemma open models and Gemini APIon-device ML and quantizationML model documentation and model cardsdemocratizing AI educationAI developer advocacyresponsible AI and bias in ML models
Timeline
15 Research15 total
2025
2024
2023
2022
2021
2019
2018
Key Contributions
Hugging Face Spaces
Launched and scaled Hugging Face Spaces from 500 to over 400,000 community-built ML demos, enabling developers and researchers worldwide to share interactive ML applications. Spaces became a cornerstone of the Hugging Face ecosystem for showcasing models and building ML portfolios.
Hugging Face Hub Growth (3k to 3M repos)
As Head of Platform and Community, led the growth of the Hugging Face Hub from 3,000 to over 3 million repositories during 3.5 years. Established key partnerships with labs (Meta, Google) and integrated open-source libraries like spaCy and Sentence Transformers.
Hands-On Generative AI with Transformers and Diffusion Models (O'Reilly)
Co-authored a comprehensive O'Reilly book (published December 2024) teaching practitioners how to use generative AI techniques to create text, images, audio, and music using transformers and diffusion models, with extensive code samples and illustrations.
Google DeepMind Developer Experience for Gemma and Gemini
Leads the global developer experience team at Google DeepMind, driving launches for Gemini 3, Gemma (180M+ downloads, 80k+ developer variants), AI Studio, MedGemma, EmbeddingGemma, and the GenAI Processors library. Focuses on empowering developers to build AI-powered products.
AI Learners Community
Co-founded AI Learners, a Spanish-speaking open community for people interested in learning about AI and its applications. Created educational ML content in Spanish and ran monthly workshops to broaden access to AI knowledge in Latin America.
The Hacker Llama (Blog & Newsletter)
Created and maintains The Hacker Llama, a personal blog and Substack newsletter covering Machine Learning, Open Source, and developer tooling. Provides weekly ML news summaries covering open model releases, new tools, and research breakthroughs.
Notable Quotes
“
You should also document your models...how it was trained, which data was used to train this model.
“
It's about transparently communicating what these models do, how they work, what could be the impacts.
“
After 3.5 years growing the open-source ecosystem at Hugging Face (from 3k to 3m repos!), it's time for a new adventure.
“
Hugging Face isn't just a company -- it's a community where kindness and collaboration thrive at its core.
12 sources(click to expand)
Omar Sanseviero -- hackerllama (personal website)osanseviero -- GitHub Profileosanseviero -- Hugging Face ProfileOmar Sanseviero on Transformer Models and Democratizing Good ML Practices -- InfoQ PodcastOmar Sanseviero -- LinkedIn (Google DeepMind)Omar Sanseviero departure post -- LinkedIn (October 2024)Hands-On Generative AI with Transformers and Diffusion Models -- O'ReillyOmar Sanseviero -- QCon Plus 2022 Speaker PageOmar Sanseviero -- Abracadata Summit 2025The Hacker Llama -- SubstackIntroducing Gemma 3n Developer Guide -- Google Developers BlogBook announcement -- Omar Sanseviero on X (December 2024)
Research generated March 19, 2026
Builders & Technical Leaders/Omar Sanseviero
All Profiles