At Mercor, I evaluate and improve agentic coding systems across SWE-bench and Terminal-bench. I create benchmark tasks from open-source pull requests and verify them in reproducible Docker environments.
At Morgan Stanley, I worked on wealth and investment management technology, including fixed-income and securities services. I designed and built hedging engines and developed an application to poll near-real-time positions from Aladdin into a DB2 database.
At Wolters Kluwer, I developed annotator services for semantic analysis, summarization, and information extraction using Flask, NLP, and LLMs. I also worked with Recognizers Text, SpaCy, and BERT to extract dates and named entities from text.
Earlier, at SymphonyAI and Macquarie, I worked on monitoring services, timeseries data, risk reporting, and regulatory reporting applications. At ALTRAN, I developed and tested Python and C++ software for real-time charging and intelligent network products.
