About
Principal Software Engineer working on AI platforms and distributed systems. Based in New York City.
My recent work is on AI agents, a semantic SQL compiler, and the tools to evaluate and observe them. Before that, I built search infrastructure and distributed job processing. I still write the code, alongside defining the work and coordinating delivery.
I joined Skubana, now Extensiv, in 2019, became a team lead in 2024, and moved into the Principal role in April 2025. Earlier, I was a full stack Java developer at Acadaca, building backend features, integrations, and operational tooling. I have a BS in Computer Science from Stony Brook University.
I own Extensiv AI, from the agent workflows to rollout and production support. The earlier Python/FastAPI and LangGraph platform reached general availability in June 2026 and included permission-scoped document retrieval. For the rewrite, I built the agent harness, observability tools, synthetic datasets, and vector-retrieved examples for multi-step reporting.
For the agentic rewrite, I built a compiler that turns the data requested by an LLM into deterministic SQL. It plans joins across 40 entities, validates business concepts, and spans SQL Server, OpenSearch, and Parquet. Coverage grew from 11 to 40 entities, with 1,000 attributes and 100 metrics. It reached 90%+ accuracy on an internal benchmark of customer questions within supported query shapes.
I wrote the PRD and engineering tickets for the rewrite and allocated delivery across four engineers. I meet with customers to review their AI workflows and refine reporting requirements. I also drove audit logging for write operations across independent product teams.
Before the AI work, I delivered an order-search migration covering 277M+ orders across 1,700 tenants. I built the backend API and DMS-based load, then owned the feature flags, rollout, and monitoring. I also worked on SQS-backed shipment processing, with Redis and WebSockets to report progress.