Corey Zumar
Software Engineer at Databricks
Berkeley, California, United States
Summary
🤩
Rockstar🎓
Top SchoolCorey Zumar is a software engineer based in Berkeley with eight years of experience building robust backend systems and ML deployment tooling at Databricks. He brings a research-minded approach from his RISE Lab background and a history of internships at Google and Airbnb, combining production-grade engineering with experimental rigor. Corey contributes to notable open-source projects — including work on Stanford's DSPy framework and MLflow’s SageMaker integration — focusing on evaluation, retrieval, and making model deployment more flexible. He has a strong testing and QA mindset demonstrated by enhancements to sqlparse’s test suite, and a practical eye for reliability shown in bug fixes and URI/compute-spec handling. Known for improving core functionality rather than flashy features, he excels at hardening systems that bridge research and production.
7 years of coding experience
3 years of employment as a software developer
Bachelor of Arts - BA, Computer Science, Bachelor of Arts - BA, Computer Science at University of California, Berkeley
French
Stackoverflow
81reputation
3kreached
1answer
1question
Github Skills (20)
continuous-deployment10
sql-parser10
docker10
pytest10
dsym10
python10
machine-learning10
mlflow10
dockers10
ml-deployment10
parsing10
data-retrieval10
evaluation-framework10
test-automation10
api-design9
Programming languages (8)
Github contributions (5)
mlflow/mlflow
Jul 2018 - Jan 2023
Open source platform for the machine learning lifecycle
Role in this project:
ML Engineer
ML EngineerContributions:13 releases, 6239 reviews, 457 commits in 4 years 7 months
Contributions summary:Corey contributed significant enhancements to the MLflow SageMaker integration, focusing on functionalities to delete and update applications deployed via SageMaker, as well as the support to add, replace, and archive different models to existing deployed applications. Furthermore, the user implemented support for specifying compute specifications like instance type and count when deploying SageMaker models, allowing for greater deployment flexibility. The user's work included modifying the PyFunc container to handle MLeap, alongside fixes and improvements related to URI handling and other general fixes across the projects.
pythonlifecyclemlmachine-learningincremental-learning
stanfordnlp/dspy
Sep 2024 - Jan 2025
DSPy: The framework for programming—not prompting—language models
Role in this project:Back-end Developer
Back-end DeveloperContributions:111 reviews, 25 PRs, 5 pushes in 4 months
Contributions summary:The user, dbczumar, primarily focused on bug fixing and improvements within the `stanfordnlp/dspy` repository. Their contributions involved changes to the `evaluate` and `retrieve` modules, including modifications to display tables and Databricks retrieval functionalities. These changes indicate a focus on refining the functionality and robustness of the core components of the DSPy framework, particularly in areas related to evaluation and data retrieval from various sources.
nlpbertknowledgepredictlanguage-models
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.