Hi, I'm Saurabh!

I am a math graduate from IIT Kanpur building production computer vision and NLP systems. My work includes co-inventing a US patent in computer vision at Adobe, publishing a research paper at KDD'21 while at Microsoft Research, and building open-source libraries. I enjoy mathematics, solo travel, fitness, and long-form reading.

Blog GitHub X (Twitter)

Work Experience

Highlights

2026Released open-source projects Mamba SSM for macOS and minFLUX; travelled to 15 countries so far — see my travel experiences.
2025Co-invented a US patent in computer vision at Adobe.
2022 - 2024Core contributor to the development and training of the audio-to-video model used in Adobe's production lipsync system. This system is used by millions of users to dub videos into multiple languages with lipsync.
2019 - 2021Co-authored a research paper published at KDD'21 while at Microsoft Research.

Open Source

Mamba SSM for macOS

Mamba State Space Models implemented for Apple Silicon GPUs.

minFLUX

Minimal, hackable implementation of FLUX diffusion models.

Get in Touch

Feel free to reach out at my_github_username@gmail.com or X (Twitter).