Summary
Shane Keller is a Staff Software Engineer in San Francisco with 12 years of experience building production ML and distributed systems across startups and large teams. Currently at Zitara, he applies embedded-to-cloud expertise to battery management software, combining ML, Go/Python services, and cloud infrastructure to accelerate decarbonization. Previously he led ML at Level, shipping low-latency fraud and underwriting models and architecting a multi-terabyte analytics platform on dbt/BigQuery. His background spans high-scale systems at Fitbit, low-latency data pipelines at a hedge fund, and open-source contributions to scikit-learn informed by statistics research. Comfortable across the stack, he blends hands-on model development (Keras, scikit-learn, XGBoost) with cloud automation (Terraform, Kubernetes, AWS/GCP) and data engineering (Spark, Dask, Parquet). He’s equally likely to optimize a production auth service for 80K RPS as to extract novel signals from time-series markets—bringing both pragmatic engineering and research-minded curiosity.
12 years of coding experience
10 years of employment as a software developer
B.A, Neuroscience, Spanish, B.A, Neuroscience, Spanish at University of Southern California
Stanford University
Portuguese, Spanish