PharmaOH
Co-Founder & Machine Learning Engineer
- Built end-to-end ML pipelines on real sports scouting data (ingest → feature store → model training → evaluation) with reproducible notebooks.
- Designed data products for pharma market-gap analysis using multi-source sales + ChEMBL data; normalized, clustered, and surfaced outliers for decision-making.
- Operationalized models with containerized services and scheduled batch inference; implemented monitoring on drift and data quality checks.
- Prototyped retrieval and labeling workflows to turn raw domain data (matches, player stats, drug indications) into high-signal training datasets.