Orchestrated Flink compute clusters for real-time streaming analytics and online ML on terabytes of daily query volume that power real-time decision making for query routing and pricing on the graph gateway (~2 billion GraphQL queries a day). Implemented custom sketches / probabilistic data structures for real time insights on network data.

Designed novel receptor proteins for controlling immune cells. Patented receptor proteins for CAR-T cell therapies.

Built ML pipelines for candidate generation, manufacturing, and optimization that incorporated multi-modal information ranging from gene expression, long read sequencing, structural biology, and cellular assays. Designed CAD software for enabling the design of therapeutic multi-domain proteins.

Implemented ML pipelines for entity resolution/record linkage and geometry (JTS) generation on terabytes of daily data ingest from metadata, polygon and mobile SDK data for maps and visits API.