Bilal Alsallakh

Join Prog.AI to see contacts

Summary

🤩

Rockstar

Bilal Alsallakh is a Senior Deep Learning Scientist based in the San Francisco Bay Area with 14 years of experience combining computer vision, model interpretability and visualization. Currently at Voxel, he previously drove ML transparency and robustness research in Meta’s Responsible AI org and exposed vulnerabilities in small-object detectors while at Bosch Research. He holds a PhD from Vienna University of Technology and pairs academic rigor with production-focused engineering. An active open-source contributor, Bilal improved PyTorch’s Captum (adding Occlusion and usability tweaks) and has contributed front-end UX enhancements to the ieeevis.org site. His work centers on using visualization not just to explain models but to actively optimize and harden them for real-world deployment.

14 years of coding experience

Github Skills (13)

html10

pytorch10

interpretation10

website-design10

front-end-development10

python9

jekyll9

image-manipulation9

machine-learning9

computer-vision9

css8

javascript7

javascripts7

Programming languages (4)

CSSHTMLJupyter NotebookPython

Github contributions (5)

ieee-vgtc/ieeevis.org

Aug 2017 - Nov 2019

Website for ieeevis.org

Role in this project:

Front-end Developer

Contributions:266 commits, 29 PRs, 265 pushes in 2 years 2 months

Contributions summary:Bilal primarily focused on updating the website's front-end elements. They added video previews to the program and poster sections. The user also updated the sidebar and added/modified images and links within the website. Furthermore, they made changes to the home page's layout, including the welcome message and photo grid.

pytorch/captum

Dec 2019 - Feb 2022

Model interpretability and understanding for PyTorch

Role in this project:

ML Engineer

Contributions:3 reviews, 82 commits, 24 PRs in 2 years 2 months

Contributions summary:Bilal primarily contributed to the development and enhancement of interpretability tools for PyTorch models within the `captum` repository. Their work included implementing new functionalities like the Occlusion algorithm, fixing existing tutorials to reflect API changes and ensuring their stability. The user also adjusted existing examples, adjusted the code to work with monochrome input and made the 'eps' parameter user adjustable in the .attribute() method. The user's commits demonstrate a focus on expanding the capabilities and usability of the Captum library.

pytorchinterpretable-aifeature-importanceunderstandinginterpretability

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.

Request Free Trial