• Prototyped novel stochastic nonlinear optimization algorithms and libraries for insurance pricing and supply chain prediction problems.

• Designed functionality for conformal uncertainty intervals on time series model predictions, based on up-to-date statistical literature.

• Architected critical portions of existing model-to-S3 infrastructure for increased system reliability and model provenance.

Developed live tracker for all busses and trains in Massachusetts Bay Transit Authority (MBTA,)

processing real-time GPS pings to calculate route headways, intuit arrivals and missing spatial

information with GTFS shape data, and track MBTA performance.

• MLE for the Document Intelligence Team, working on problems related to document understanding from R&D to production.

• Iterating on ML and engineering solutions for document layout analysis, information extraction, and OCR, with a focus on applying GNN's to this domain. (Pytorch)

• Maintained the ML model lifecycle for a fully productionized key-value extraction system, which has processed over 2 million broker reports to date. (Airflow, Python, SQL)

• API Services Engineer, productionizing ML financial products, including our broker research document information extraction API (Python, Kafka, Flask, Prometheus)

• Search Software Engineer, developing search engines for S&P's Market Intelligence next-generation platform and Kensho's financial documents platform, Codex. Patent pending. (Elasticsearch, Python)

• Graph Infrastructure Engineer for Kensho Technologies' financial events knowledge graph (OrientDB, GraphQL, Python)

• Head Data Structures TA, 1 year

• Algorithms CA, 1 year

• Data Structures CA, 1.5 years