Natasha Kononenko
Staff Software Engineer at Riot Games
Los Angeles Metropolitan Area United States
Summary
🤩
Rockstar🎓
Top SchoolNatasha Kononenko is a Staff Software Engineer in the Los Angeles area with six years of hands-on experience building production-grade systems and ML compiler tooling. Her background at Google and DeepMind focused on lowering and optimizing ML operations—contributions that include casting and lowering fixes in TensorFlow/TFLite to TOSA and implementation of operator support in high-performance projects like iree-org/iree. She blends backend systems engineering with machine-learning compiler work, shipping end-to-end tests and conversions for nontrivial ops such as grouped convolution, bucketize, and string concatenation in compiler runtimes. Known for pragmatic bug fixes that unlock real functionality, she moves models closer to efficient deployment while keeping code maintainable. A Worcester Polytechnic alum with a game-development foundation, she brings an unusual mix of systems-level rigor and product-minded practicality to ML infrastructure problems.
6 years of coding experience
7 years of employment as a software developer
Bachelor of Science (B.S.), Computer Science, Interactive Media and Game Development, Bachelor of Science (B.S.), Computer Science, Interactive Media and Game Development at Worcester Polytechnic Institute
English, Hebrew, Russian
Github Skills (15)
tflite10
machine-learning10
togaf10
c-language10
lsa10
tensorflow10
cprogramming-language10
tosca10
mlr10
compiler-development10
python9
testing9
neural-network9
deep-learning9
compiler-construction8
Programming languages (5)
Github contributions (5)
iree-org/iree
May 2020 - Sep 2022
A retargetable MLIR-based machine learning compiler and runtime toolkit.
Role in this project:
Back-end Developer
Back-end DeveloperContributions:65 reviews, 100 commits, 202 PRs in 2 years 4 months
Contributions summary:Natasha implemented and tested string concatenation functionality within the "iree-org/iree" repository, specifically for the strings module. Their work included implementing the "Concat" function, adding related end-to-end tests, and modifying various files. Additionally, they added support for the "mhlo.exponential_minus_one" operation, further contributing to the compiler's capabilities. They also added support for the "mhlo.not" operation, which included conversion from HLO to VMLA and tests.
mlirspirvvulkantensorflowcompiler
tensorflow/tensorflow
May 2020 - Jan 2023
An Open Source Machine Learning Framework for Everyone
Role in this project:ML Engineer
ML EngineerContributions:2 reviews, 7 commits, 3 comments in 2 years 7 months
Contributions summary:Natasha contributed to the TensorFlow repository by implementing and refining the lowering of TFLite operations to the TOSA (Tensor Operator Set Architecture) dialect. Their work focused on handling data type conversions, specifically casting i64 indices to i32 for gather operations, and supporting negative dimensions in argmax operations. Furthermore, the user added lowering for bucketize, which involves broadcasting the input and boundaries and performing comparisons. They also addressed the grouped convolution, including bug fixes and new features to make it work correctly.
pythondata-sciencedeep-learningmlmachine-learning
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.