Built an AI-powered debugging platform used by 100+ companies, where frontend errors and user events are captured via a Sentry-based script and routed through a custom FastAPI proxy to Kafka for asynchronous processing.
Contributed to the ingestion service (Node.js) that consumes Kafka streams, processes raw error data, enriches it with AI-generated insights, and stores it in PostgreSQL, enabling real-time debugging and analytics.
Worked on the Exports App within Smarsh's enterprise-grade data archival platform, enabling financial institutions to request, schedule, and retrieve filtered data exports at scale.
Developed frontend components (JavaScript) that allowed users to initiate exports with custom filters and track progress in real time, while coordinating with backend services written in Java to queue and process export jobs via Kafka.