Tal Ashuach
Head Of ML-omics And Computational Biology
San Francisco, California, United States
Summary
🤩
Rockstar🎓
Top SchoolTal Ashuach is an Associate Director leading ML-omics and Computational Biology at insitro with eight years of experience bridging machine learning and single-cell/omics research. He holds a PhD in Computational Biology from UC Berkeley and has progressed rapidly through research and leadership roles across startups and academia, including Vevo Therapeutics and Patch Biosciences. Tal contributes to prominent open-source work in single-cell analysis—notably scvi-tools—where he implemented and tested improvements to the PeakVI probabilistic model. He combines hands-on model design, robust testing, and production-focused engineering to translate complex biological data into actionable insights. Known for pragmatic architecture changes (e.g., encoder modularization) and attention to imputation and injection mechanisms, he excels at making advanced probabilistic methods reproducible and deployable. Based in San Francisco, he pairs deep domain expertise with a track record of scaling ML-omics efforts from research prototypes to team-led programs.
8 years of coding experience
9 years of employment as a software developer
Doctor of Philosophy (PhD) Computational Biology, Doctor of Philosophy (PhD) Computational Biology at University of California, BerkeleyBachelor of Science (B.Sc.) Computational Biology, Bachelor of Science (B.Sc.) Computational Biology at The Hebrew University of Jerusalem
English, Hebrew
Github Skills (8)
rna-seq10
seq10
pytorch10
deeplearning-ai10
deep-learning10
python10
sc10
machine-learning8
Programming languages (4)
Github contributions (5)
scverse/scvi-tools
Jan 2021 - Oct 2021
Deep probabilistic analysis of single-cell and spatial omics data
Role in this project:
Data Scientist
Data ScientistContributions:10 reviews, 54 commits, 7 PRs in 8 months
Contributions summary:Tal primarily contributed to the scvi-tools repository by implementing and refining features related to the PeakVI model. Their work included modifying the model's architecture (separating the Encoder class, extending deep injection support), fixing bugs in the imputation method, and adding testing infrastructure for the PeakVI model. Furthermore, the user made code formatting improvements and addressed minor bug fixes within the PeakVI codebase.
hierarchicaldeep-generative-modelmixture-of-expertssingle-cell-rna-seqsingle-cell
YosefLab/MPRAnalyze
Nov 2017 - Nov 2020
Analyse enhancer strength based on MPRA experiments.
Contributions:164 commits, 1 PR, 44 pushes in 3 years
enhancerstrength
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.