I build practical software that connects real systems and keeps AI output reviewable.

My portfolio focuses on neutral, from-scratch demos: serverless agents, LLM tool use, integrations, persistence, analytics interfaces, and human-in-the-loop review surfaces. The model proposes, code validates, and a person stays in control.

Serverless agents · LLM tool use · integrations · analytics · grounded retrieval

Selected Work

Three invented-domain projects showing production-shaped engineering without leaning on private or sensitive domain context. See the full project index →

Profile

I'm Greg Finin. I like building useful software at the boundary between messy real-world records, operating decisions, and systems that need to be trusted before they are automated.

The portfolio projects here are intentionally neutral and sample-data-only. They are built to show engineering judgment: clear data contracts, persistence, tests, deployable service shapes, source citations, and review states.

The through-line is simple: connect the records, make the system behavior inspectable, validate structured outputs in code, and keep a human review path wherever AI is involved.