Jason Sleight
Principal Software Engineer at Yelp
Northville, Michigan, United States
Summary
🤩
Rockstar🎓
Top SchoolJason Sleight is a Principal Software Engineer based in Northville, Michigan, who blends a PhD-level grounding in AI with hands-on systems and ML engineering to drive enterprise-grade ML platforms. At Yelp he shapes AI strategy with executives, builds core MLOps capabilities, and led the development of centralized feature stores, model serving, Spark infrastructure and Jupyter-backed workflows that power production ML. His open-source contributions—such as improving Spark integration in Yelp's paasta and adding XGBoost parameter serialization to MLeap—reflect a focus on reliable model deployment and cloud interoperability. With a decade-plus academic-to-industry trajectory converging research rigor and production pragmatism, he is equally comfortable designing high-level AI roadmaps and shipping low-level serialization and credential-management fixes. Colleagues rely on him for bridging product, legal, and engineering concerns around responsible AI adoption.
5 years of coding experience
7 years of employment as a software developer
Doctor of Philosophy - PhD Artificial Intelligence, Doctor of Philosophy - PhD Artificial Intelligence at University of Michigan
Github Skills (24)
kubernetes10
xgboost10
spark10
python10
machine-learning10
data-serialization10
serialization10
scala10
aws10
kubernetes-pods10
command-line-interface9
data-pipelines9
command-line9
cli9
data-pipeline9
Programming languages (4)
Github contributions (5)
combust/mleap
Nov 2020 - Jan 2023
MLeap: Deploy ML Pipelines to Production
Role in this project:
Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:4 releases, 85 reviews, 40 commits in 2 years 2 months
Contributions summary:Jason primarily contributed to the serialization of additional parameters for the XGBoost models within the MLeap framework. This involved modifying code related to loading and saving XGBoost models to ensure all necessary parameters were correctly handled during the serialization and deserialization processes. Furthermore, the user updated parity tests and made modifications to support the serialization of various XGBoost parameters. These changes directly impact the deployment of machine learning pipelines by enabling model persistence and ensuring consistent behavior across different environments.
transformerspythonml-pipelinesmleapdata-science
Yelp/paasta
Jun 2022 - Oct 2022
An open, distributed platform as a service
Role in this project:Back-end & DevOps Engineer
Back-end & DevOps EngineerContributions:14 reviews, 13 commits, 10 PRs in 3 months
Contributions summary:Jason primarily contributed to the `spark-run` command within the `paasta_tools` project. Their work involved adding and modifying arguments related to AWS credential handling and temporary credential providers for Spark. They also addressed issues related to cluster managers (Kubernetes/Local). These changes indicate a focus on improving the integration of Spark applications with cloud environments and providing users with more flexibility in managing their deployments.
mesosdockerplatform-as-a-servicedistributed-systemsmicroservices
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.