Tony Chen

Contributed to the development of an internal LLM-powered enterprise knowledge assistant (powered by Amazon Bedrock) enabling engineers to retrieve operational knowledge and documentation via natural language queries. Designed and implemented retrieval-augmented generation (RAG) pipelines using internal data sources to fetch contextual enterprise information. Built vector retrieval pipelines with embeddings for semantic search across engineering documentation and operational metadata. Developed backend services using Python, FastAPI, and PostgreSQL to orchestrate knowledge retrieval workflows and deliver AI-generated responses to internal systems. Implemented conversation orchestration logic supporting multi-tenant enterprise knowledge queries across internal documentation and engineering resources. Built ML experimentation pipelines in PyTorch to analyze operational telemetry patterns and explore ranking signals for knowledge retrieval. Integrated enterprise data sources including int

Worked on an internal service telemetry, incident detection and response platform. Built dashboards, backend services, and REST APIs to visualize and monitor service health metrics, alerts, and operational events in real time. Optimized API performance and database queries to improve dashboard responsiveness. Collaborated with multiple infra and engineering teams to integrate operational monitoring with existing AWS infrastructure and tooling.

Developed a full-stack internal dashboard used by engineering teams to monitor service health, investigate operational anomalies, and analyze service activity. Built frontend components using React.js and TypeScript, and backend services with Java and Python, connected to PostgreSQL for incident metadata and service lifecycle data. Focused on API performance, data modeling, and dashboards that support real-time telemetry visualization.