I help restaurant operators and business owners find the real problem underneath the one they're trying to solve, then build something that actually fixes it.
The work always starts in the same place: what decision are you actually trying to make, and what's standing in the way? The answer might be an AI-powered tool, a data system, or a reframe of the problem before anything gets built. It depends on what the diagnosis finds.
For nearly 30 years, I managed operations in large-scale food service: hundreds of locations, thousands of menu items, margins that don't forgive mistakes. The tools available were never built for the decisions I was actually trying to make. So I learned to build my own.
When AI made it possible for someone with an operations background, not a development background, to build real software, I didn't start with the technology. I started with the problems I'd been staring at for three decades. Starting with the problem changes what you build.
When you start with the problem, you build smaller things. More specific things. Systems that keep running long after the project closes, because they were built around the decision that needed to be made, not the technology that was available.
Most operators who reach out think they have a data problem or a software problem. Sometimes they're right. More often, what they're pointing at is a symptom and the real problem is somewhere upstream.
A purchasing operation that looks disorganized is often a visibility problem. A software platform nobody uses is usually a trust problem, not a training problem. A validation gap that's been deferred for years can often be closed in weeks with the right AI tool. The pattern repeats: the surface problem points to a different root cause, and solving the wrong one just moves the cost around.
I've built solutions across all of these. What they have in common is that they started with a problem diagnosis, not a technology choice. The tool came second. That's why they last.
I don't have a methodology deck to show you. What I have is a track record of building things that kept working long after the project closed.
Buying decisions were happening without a clear picture of what was actually being used, what things cost across the menu, or whether a change in one ingredient was quietly affecting margin everywhere else.
Built a system that connected recipes, purchasing, and costs so a change in an ingredient shows up where it matters, before it shows up in the P&L.
Food cost became something that could be managed, not something discovered after the fact. Purchasing got more consistent. A major spend that was moving toward approval turned out to be unnecessary once the real picture existed. The platform ran for over a decade.
A system had been purchased and rolled out. Most of the people it was built for weren't using it. Training had been tried. It hadn't worked. The tool was quietly failing while the problems it was supposed to solve kept getting worse.
Instead of another training push, sat with the people doing the work and found out exactly what was slowing them down. Redesigned the process and education around those specific friction points, not the feature list.
When the tool started saving time instead of costing it, adoption took care of itself. Recipe errors came down. Inventory variance came down. Teams that had resisted started using it. Teams that hadn't been approached started requesting it.
There was a category of work that everyone knew needed to happen: reviewing data quality, catching errors before they compounded. It never got done because nobody had time to do it consistently.
Used AI to design a self-service audit process and deliver a working platform in three weeks, with documentation so the team could run it, expand it, and own it without coming back.
The team moved from working around data they weren't sure they could trust to actually knowing. Decisions that had required manual cross-checks became straightforward. The review work that had been deferred for years became part of the normal routine.
A significant purchase was moving toward approval. The justification was based on what people believed they had across locations, not what was actually there. Nobody had a clear picture of current inventory, so the assumption was that more was needed.
Built a visibility system that gave all locations a shared view of what existed and let them lend and borrow across sites. The real picture came together quickly.
The purchase decision changed. Not because someone said no, but because the information finally existed to make it correctly. What looked like a shortage turned out to be a distribution problem. The spend was avoided before a single order was placed.
Every engagement I take follows the same three steps.
A 30–45 minute conversation on Google Meet. I want to understand the problem: not the symptom you're managing, but the decision you're trying to make and what's standing in the way. No pitch, no deck. Just a direct conversation about what's broken and what you've already tried.
Free. If it's not a fit, I'll tell you directly. If I know someone better suited, I'll say so.
If there's a fit, I conduct a structured assessment of the real problem. I review what data, systems, and processes exist, identify where the gap actually lives, and deliver a clear written recommendation: what the solution is, what it will take to build, and a frank assessment of whether I'm the right person to build it.
Flat fee, credited toward the build if you proceed. Useful on its own whether or not we work together further.
Scope, timeline, and price are defined by what the diagnostic found, not by a service menu. I build the tool, system, or process that solves the real problem. Designed to run without me. Documented so your team owns it. Built to expand without starting over.
The measure of whether it worked is simple: is it still running two years from now?
I take on a small number of engagements at a time. The ones I take, I finish.
If something in here sounds like the conversation you've been trying to have, reach out. Describe the problem, not the solution you think you need. Just tell me what's broken and what it's costing you. I'll tell you honestly whether I think I can help.
I'm currently taking on a small number of new engagements.