Tiferet Gazit
Staff Machine Learning Engineer at GitHub
San Francisco Bay Area United States
Summary
🤩
Rockstar🎓
Top SchoolTiferet Gazit is a Staff Machine Learning Engineer in the San Francisco Bay Area with a decade of experience building deep learning, NLP, and graph neural solutions for code understanding and security. At GitHub she develops semantic models and GNNs that power vulnerability detection and developer tooling, contributing to high-profile projects like CodeQL with work on adaptive threat modeling and sink classification. Her background in physics, math, and medical computer vision (BS from Brown, MS from MIT) gives her a strong foundation in principled modeling and cross-domain problem solving. She combines research-caliber thinking with production engineering to translate complex program analysis into deployable ML pipelines. Colleagues rely on her ability to refactor and generalize detection logic, improving training data extraction and feature engineering for robust security models. Outside core ML tasks she studies the societal impact of LLMs and intelligent algorithms, aiming to shape safer, more useful AI for developers.
10 years of coding experience
MS, Computer Science / Medical Computer Vision, MS, Computer Science / Medical Computer Vision at Massachusetts Institute of TechnologyBS, Physics, Math, BS, Physics, Math at Brown UniversityWeizmann Institute of Science
java, c++, python
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (6)
dataflow10
codeql10
security9
machine-learning9
it-security9
javascript8
Programming languages (8)
Github contributions (5)
github/codeql
Nov 2021 - Jan 2023
CodeQL: the libraries and queries that power security researchers around the world, as well as code scanning in GitHub Advanced Security
Role in this project:
ML Engineer
ML EngineerContributions:94 reviews, 229 commits, 32 PRs in 1 year 2 months
Contributions summary:Tiferet's commits primarily focus on developing and refining components related to adaptive threat modeling within the CodeQL framework. Their work includes creating and modifying code to classify and categorize endpoints, specifically focusing on identifying and characterizing sink types. They have generalized the definition of known sinks and refactored code, demonstrating an understanding of data flow analysis and model building. The user has also implemented new filtering and characteristic features for the training data extraction process.
github-advanced-securitysecurity-toolsgithub-security-labsecurity-researcherssecurity
tiferet/keras
Dec 2017 - Jan 2019
Deep Learning library for Python. Convnets, recurrent neural networks, and more. Runs on Theano or TensorFlow.
Contributions:24 pushes, 2 branches in 1 year
pythondeep-learningrecurrent-neural-networkskeras-neural-networkstheano
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.