Hi, I'm Michał.

I connect business priorities to the engineering needed to make AI useful in practice. That means understanding the workflow, choosing where AI adds value, and building a system that a team can operate. I favor the simplest approach that will hold up.

I split my time between technical leadership and the code itself. At Google, I led an AI support-agent project that prepared review-ready responses for more than 70% of technical customer inquiries. I also led architecture and integrations for products used at scale. Today I lead engineering at Delta Labs, working on the multi-agent architecture, validation, and evaluation behind Elaiia.

Before that, I co-founded Notamify and served as its CTO. We turned raw NOTAM data into structured pre-flight briefings and eventually sold the company. Through Practically Agents, I help teams discover, deliver, and improve AI systems. I also write about the decisions behind production AI. If you’re considering a project, discuss an AI transformation with me.