Data science that ships, in whatever domain the problem lives.

Senior Data Scientist

• Forecast thermal-runaway risk and warranty life for 10+ vendor and in-house battery cell programs, combining lab stress tests under 98th-percentile fleet conditions with 1 TB of telemetry from 3.5M vehicles. As cell owner, I sign off on go/no-go qualification decisions.
• Own the final quality gate for every newly manufactured Tesla 4680 battery before it enters a vehicle. I develop and deploy defect detection models that segment 50+ battery components in CT scans, delivering 98% recall on critical checks.
• Saved 200+ technician hours a week by building a quality-analytics pipeline and live dashboards used daily by 45 engineers across 4 teams. The pipeline inspects 2,000+ batteries a day via Celery/Redis on 6 VMs and flags out-of-spec cell metrology through Splunk alerts.

Data Scientist

• Led product-placement A/B tests for Compass Group USA, randomized by store and week, lifting interaction-to-purchase conversion from 24% to 31% (+29%).
• Partnered with product and sales to define shopper engagement metrics (funnels, cohorts, repeat visits) and built an in-house Python library to compute them in Databricks and Snowflake, winning $600K in new contracts over 3 months.
• Built a Monte Carlo simulation of shopper paths and shelf-to-shelf travel times to predict how store layout changes shift shopping behavior. The product-placement swaps I recommended to retail clients lifted revenue 15% across two stores.

Cell Engineer Intern

• Improved defect-detection recall by 40% over the previous binary filter with a Vision Transformer model on CT scans, catching core impingement and collapse before cells moved downstream.
• Replaced several $27K Keyence machines with a CT-based method that non-invasively measures terminal/lid bulge on cells to within 50 µm, so engineers inspect cells without new equipment.
• Built a daily-refreshing Tableau dashboard linking dissection reports to production metrics. It surfaced failure trends that drove electrolyte loading and cell design changes, cutting electrolyte leaks by 20% in 5 months.

Summer Associate, Credit Fraud Risk

• Improved small-business lending decisions by lifting the PySpark loan default model 14%, with features for erratic repayment, seasonal revenue shifts, and credit-utilization spikes.
• Made high-risk applicant segments explainable to risk stakeholders and federal regulatory auditors using feature binning and hierarchical clustering, giving clear reasons for each segment.

Undergraduate Researcher

• Built a data pipeline for the optimal compressor wash interval on AE3007 engines and surfaced it in an R Shiny tool with geospatial views of hundreds of at-risk aircraft, cutting the time for maintenance planners to isolate flights needing a wash from 3–4 hours to under 5 minutes.

Research & internships