Ibuildsystemsthatship.
Full-stack products and AI automation for teams who'd rather see it working than hear about it.
Three systems, three constraints.
Most automation ideas die in a slide deck.
The gap is never the idea, it's whoever has to actually wire it into a pipeline that runs unattended. That's the part I build.
Support tickets pile up faster than anyone can read them. By the time a pattern is noticed, it's already cost a week.
A RAG pipeline ingests new tickets on a cron job, classifies them, and surfaces what's actually going on underneath the noise.
Insights are ready before the standup, built and shipped as an engineering intern at ConnectWise.
Same principle at EaseworkAI: agentic Crew AI workers handled supplier-buyer communication end to end, over the Graph API, with no one relaying messages by hand.
What that looks like in practice.
AI Systems
Turn unstructured data into decisions someone can actually act on.
Full-Stack Products
Ship the product, not just the demo that works on one machine.
Automation
Remove the manual step that only exists because no one automated it yet.