Designed and implemented RAG-based AI virtual shopping assistant and livestream copilot, boosting engagement and conversion by answering shoppers' questions.
•
Optimized event-tracking data ingestion pipelines, reducing data transfer and storage costs and saving ~$110K per year.
•
Architected and developed conversational analytics pipeline enabling merchandisers to extract insights (e.g., trending questions, product sentiment).
•
Instrumented distributed tracing and LLM observability tools to monitor errors, optimize system latency, reduce cost, and improve LLM response quality.
•
Optimized CI/CD pipelines, GitHub Actions, and QA/release processes, reducing production bugs and enhancing developer productivity by over 24%.
•
Improved time-to-resolution by 50% by authoring runbooks and streamlining incident response process
•
Mentored junior engineers and collaborated cross-functionally with Product, Design, and QA teams.
