I joined Clear Street in January 2026 as a software engineer on clearing and settlement, working mostly on corporate actions: dividends, stock splits, mergers, and the platform that processes them for a prime broker. The work is Go services on Kafka and Postgres, where correctness and ordering matter as much as throughput.
Custody is a small team with a lot of surface area, so the work is a mix of new features and the performance and reliability behind them, and I own most of what I ship end to end, from design through production. On the performance side, I led the move from a five-minute batch refresh of voluntary corporate actions to real-time processing, which does far fewer jobs for the same result and cut intraday database load by roughly 80%. I also optimized the job-running framework to take the same daily window from 445K to 1.6M jobs, which removed a month-end bottleneck, and reduced database and CPU pressure on the hot services behind that path so ledger ingestion went from 40K to 90K events per second.
Alongside that, I built the observability the team runs on: shared metrics libraries, Kafka and service dashboards, panic detection, and end-to-end job telemetry, plus workflow-duration alerting after finding a production workflow that had stalled without raising an error.
Outside of work I contribute performance changes upstream to Go. See Go for the list.