I don't just design the system. I build it.
That's what makes this practice different from every strategy deck and vendor pitch you've sat through.
Operator first. Builder second. Consultant third.
Most failed AI projects start with a tool instead of a problem. A vendor demos something impressive, leadership signs off, and six months later there's a half-adopted platform that never touched the thing that was actually broken.
I started this practice because I kept watching that happen, and the order was backwards every time. Understand the business first. Then decide whether AI belongs anywhere near it.
My background is operations, not AI research. I spent nearly a decade at Zendesk, a $2B SaaS company, running programs across product, strategy, and go-to-market. Before that, Big Four consulting and an MBA. That decade taught me that the hard part of any system is rarely the technology. It's getting a team to actually use the thing, keep it running, and treat it as theirs.
The unusual part: I build, too. Inside companies, that has meant agentic workflows, retrieval systems, and getting a stack of disconnected tools to actually talk to each other. The unglamorous transformation work that makes AI stick. On my own, I've shipped live AI applications with no development team, and they're running in production today with real users. When I recommend something in an engagement, it's because I've built it or something close to it. I won't hand you an idea I couldn't execute myself.
And sometimes the honest answer is that AI isn't your problem. The data is a mess, or a process is broken, or two teams haven't talked in a year. I'll say so. Solving the wrong thing with the right technology is still the wrong thing, and you're not paying me to be agreeable.
The track record behind the work
The principles behind every engagement
AI isn't a tool upgrade. It's a structural shift.
Companies treating it like a software rollout are going to fall behind. Real transformation means aligning your strategy and your goals with how work actually gets done. Not just adding a new layer on top of a broken process.
The real problem is usually one level up from where everyone is looking.
Clients come in thinking they need a tool, a new workflow, or better reporting. Sometimes they do. More often, the issue is upstream. Finding that before touching anything else is the actual work.
Most transformations fail at rollout, not at design.
A solution that isn't easy to adopt just becomes another thing people work around. I build for repeatability from the start. A system that only works when someone is explaining it isn't a system yet.
Change that doesn't bring people with it doesn't stick.
AI is changing what work looks like. Sometimes the goal is doing more with fewer people. I don't pretend that conversation isn't happening. I help you move through it honestly, because change that bypasses your team doesn't hold.