Led backend and data platform development for investment research, market insights, and ranking pipelines.

• StackPython, FastAPI, LangChain, AWS (SNS/SQS, ECS, Lambda, Elastic Beanstalk)
• DataClickHouse, Snowflake, Airflow

Selected Work

• Built event-driven architecture using SNS, SQS, and FastAPI, based on a pub/sub pattern, to support market insights and ranking pipelines.
• Designed GenAI modules with LangChain for multi-step prompt workflows, structured JSON outputs from LLM responses, and robust text chunking for downstream processing.
• Developed high-volume async consumers that parsed emails and newsletters since 2016, extracted structured content, persisted results to domain models, and identified high-performing companies via LLM-powered entity extraction and scoring.
• Refactored legacy data pipelines into Airflow DAGs and ECS-based services for scalable ingestion and enrichment across multiple data sources.

Impact

• Processed more than 2,000 emails per day through realtime ingestion services.
• Built LLM-powered extraction and scoring workflows that contributed to profitable investment decisions by surfacing stronger company signals.
• Helped investors evaluate companies with less manual research and better ranking signals.