Tam Edwin

I've built AI-enabled operational workflows for Hado Real Estate Company, A&C Construction Company, and Grace Travels, connecting REST APIs, databases, structured AI outputs, and human-review checkpoints.

At Grace Travels, I created an n8n and OpenAI email-intake workflow that classified enquiries, routed them correctly, logged activity, and triggered follow-up. At A&C Construction Company, I built meeting-intelligence workflows that turned transcripts into summaries, action items, owners, and follow-up communication.

I built RailGuard AI, a predictive-maintenance project processing 1,516,948 time-series sensor records into 252,720 model-ready windows with 40 features. I developed the Python and scikit-learn pipeline and deployed a FastAPI service in Docker on AWS EC2 with Nginx, IAM-based S3 access, and a security review before release.

I'm comfortable designing n8n, OpenAI, RAG, multi-agent, and human-in-the-loop systems, troubleshooting integrations, and documenting workflows so teams can maintain and reuse them.