Role overview

Software Engineer responsible for designing and operating cloud-native systems on GCP. Working across data platform architecture, event-driven ETL, and production GenAI workflows in a fast-moving team.

Key contributions

• Designed a multi-tenant data platform on GCP (Cloud Run + BigQuery) supporting 5 isolated tenants and ~200K records/day with zero-downtime deploys.
• Built an event-driven ETL pipeline handling 1M+ events/day, using idempotent consumers and automated retries to minimize duplicate processing.
• Developed production LLM pipelines on Claude Opus/Sonnet — structured prompt chains, agent loops, and output validation — deployed as live internal automation.
• Shipped REST APIs with <500ms p99 latency under concurrent load; added RBAC, rate limiting, and middleware to support a 99.9% data-integrity SLA.

Technologies used

Google Cloud Platform (Cloud Run, BigQuery, Pub/Sub, IAM, GCS), FastAPI, Docker, Claude Opus/Sonnet, OpenAI APIs, Redis, event-driven microservices.