I'm Lucas. For about thirteen years I have done full-stack development and quality assurance — from the first draft through the release to everything that shows up in operation afterwards.
These days I mostly build and run agentic AI systems: the loops, tool integrations and guardrails underneath the assistants people actually use.
What interests me most is the road from demo to production — proper evals, honest QA and systems that hold up under real conditions.
In this blog I write down what I learn along the way — mostly about agents, evals and production-ready AI.
Contact
Questions or remarks? Write me an e-mail: plattcode@tuta.io
