About
Twenty years of delivery. Building with AI since 2015.
I'm Jesse Burcsik. My career has been building software, leading teams, running products and delivering public services at scale. I have been in the AI industry since 2015, starting on the founding team of an AI-first company.
How I work
Most of the work in an AI system is fitting it into the systems an organization already runs, handling the cases where it is wrong, keeping it secure, and getting people to trust it enough to use it. Those are delivery problems, and delivery is what I have done for most of my career.
My current work includes agents on private servers, voice agents, language-model data pipelines, the security around them, and auditing what AI assistants say about businesses and why. Some of it is on the work page. Other client work, built with a partner team, is not mine to show.
I write up what I find on Insights: tools and techniques I have tested on real work before recommending them.
What I actually think
Build the smallest useful thing first
The big AI transformation project is how budgets get eaten with nothing to show. One process, working properly, in production, beats a roadmap every time.
Mandates do not create adoption
People adopt what they have been shown on their own work, with clear rules about what is safe.
You should be able to fire me
I hand over the code, the accounts and the knowledge. If keeping you dependent is the business model, the work was never that good.
Boring beats clever
The simplest thing that works, every time. Impressive demos and systems that survive a year are usually not the same systems.
Agree the scope early
Every project is scoped on a short call. You get the scope and the timeline before the work starts, and you hear about a problem the week it happens.
Want to talk about it?
Tell me what you want from AI, or what you are worried AI says about you. Half an hour, no pitch, and I will tell you if I am not the right person.