Kirill Petrov
Старший Инженер По Разработке at Intel
Nizhny Novgorod, Nizhny Novgorod Oblast, Russia
Summary
🤩
RockstarKirill Petrov is a senior software engineer at Intel with a decade of hands-on experience in backend development and data-oriented optimization. He contributes to high-performance open-source projects like oneDAL and scikit-learn-intelex, focusing on performance regressions, data layout issues, and tooling automation. His work shows deep practical knowledge of numerical libraries and Intel tooling (MKL-FPK, instrumentation/tracing) as well as careful handling of Pandas-to-NumPy interactions for ML pipelines. Colleagues rely on him to bridge low-level performance fixes with reliable test coverage across diverse array formats and heterogeneous dtypes. Based in Nizhny Novgorod, he combines production-grade engineering at a major CPU vendor with active contributions that accelerate common ML workloads. A not-obvious strength is his dual fluency in scripting release automation and in writing detailed tests that prevent subtle performance regressions.
10 years of coding experience
Github Skills (21)
python10
scikit10
pandas10
machine-learning10
numpy10
performance-optimization10
scikit-learn10
data-analysis10
c-language9
cprogramming-language9
shell8
batch8
sh8
data-structures8
data-structure8
Programming languages (7)
Github contributions (5)
uxlfoundation/scikit-learn-intelex
Mar 2020 - Jul 2021
Extension for Scikit-learn is a seamless way to speed up your Scikit-learn application
Role in this project:
Data Scientist
Data ScientistContributions:9 releases, 540 reviews, 120 commits in 1 year 4 months
Contributions summary:Kirill's contributions primarily focused on improving the performance and addressing issues related to the integration of Pandas DataFrames with the Scikit-learn-intelex library, an extension for accelerating Scikit-learn. The commits involve significant work on testing and fixing performance regressions, including adding new tests for various DataFrame formats (C-contiguous, F-contiguous, non-contiguous) and data types, specifically focusing on the `to_numpy` method and its interaction with the library's underlying data processing. The user also addressed issues related to heterogeneous data types and multi-dtype arrays, demonstrating expertise in data manipulation and optimization within the Scikit-learn-intelex framework.
scikit-learnoneapimachine-learning-algorithmsdata-analysismachine-learning
uxlfoundation/oneDAL
Sep 2019 - Aug 2021
oneAPI Data Analytics Library (oneDAL)
Role in this project:Backend Developer
Backend DeveloperContributions:3 releases, 379 reviews, 110 commits in 1 year 11 months
Contributions summary:Kirill's commits primarily focus on modifying scripts related to the Intel MKL-FPK package within the oneDAL repository. These changes involve updating URLs, versions, and file paths within the scripts to align with new releases and updates of the Intel daal/daal-fpk repository. These updates include changes to batch files and shell scripts to download and extract the MKL-FPK package. Additionally, the user added profiler information to the core algorithms using Intel's instrumentation and tracing technology.
data-analyticsoneapionedaldata-analysismachine-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.