I set up AI inside your company. Then I teach your team to run it.
Bringing AI into a team either unlocks what your people can do, or turns into a giant mess. I help teams do it the first way: simple to use, easy to keep running, and explained in a way that even the least technical person can use it. And I stay until your people can run it without me.
"I expected something complicated. Thomas made it simple, and it was done so quickly. The team couldn't believe it."
"Before Thomas helped us, I felt intimidated by AI and like I was falling behind. He explained it in a way that made sense and built systems I now use every day."
"I'm not super technical, but the way Thomas explained things and set up our system made it super easy to use."
Putting AI into a team is part technical and part people.
The first half is making it exist. The second half is making it stick, and that is where most AI projects quietly stop working. I do both, and I'm the only person you deal with.
Make it exist
I build the thing.
The process, the tools, the data, the thing that runs. I build inside your tools, on your data. No new platform to learn, nothing to migrate.
- I've automated processes in every company I've worked in, long before AI.
- Two years building software with AI every day, eight apps shipped.
- Most recently, a 25-person agency's whole team set up on one shared system: shared processes, approvals in their chat tool, and a hub that shows who built what.
Make it stick
Building it is the easy part. Getting a team to actually use it is the work.
Fear of being replaced. Habits that don't want to move. Nobody wanting to ask for help. What actually holds people back:
- Being comfortable with the technology.
- Understanding what it's actually doing.
- Thinking like a data person. Not everyone is, and not everyone has to be.
- Change itself, and the question underneath it: "Am I going to lose my job to this?"
Outside of consulting I'm a mindset coach, five years one-on-one and in groups. I know how to sit with someone who's afraid of a change and get them moving. This is the conversation I'm good at.
Two recent engagements
Both anonymous at the client's request. Both still running today.
A whole agency working with AI, not just the two people who were curious.
Every teammate is set up with one command, on Mac or Windows. Approvals happen in the chat tool they already use, and a shared hub shows who built what, so the work spreads instead of sitting with two people.
- Before
- Two teams on two systems. Client reporting done by hand every week. Two people using AI on their own.
- After
- 25 people on one shared setup. 16 processes written down and shared. Reporting runs itself and the team reviews it.
A sales pipeline the founder runs directly, without writing a line of code.
Signups, follow-ups, bookings and call notes, handled on the tools the team already had. The founder set it up from the process we wrote down together, and changes it when the business changes.
- Before
- The founder chasing signups, follow-ups and bookings by hand, between everything else, and paying for five different tools to do it.
- After
- The pipeline runs three times a day and drafts every email. The founder reviews, sends, and changes it directly.
My process
One process, one team, a few weeks. Then the next one.
Define the process
I sit with your team and write down how the work really gets done: the steps, the judgement calls, and what good looks like. Clearly enough that AI could follow it.
Build it
I set up the AI to run that process on the tools you already use, and test it on one real project before it goes any further.
Hand it over
I set your people up one at a time and train them to run it and change it themselves, with me next to them until it's a habit, not a tool.
Keep it working
Tools and models change. Pick a support plan: I keep what we built running, or I keep it running and keep improving it, automating new things as they come up.
Pricing is customized for every client I work with.
AI work is different for every team: different tools, different processes, and a lot of small one-off details. After we meet, I'll send you a customized pricing sheet with the different options that make the most sense for your team and what you want to do with AI.
Ways to work together
Two different things, depending on where you are. If you know the one process you want off your plate, I build that automation. If you want the whole team working with AI, I build one shared system everyone in it uses. Not sure which? Start with the sprint: it rolls straight into whichever you pick. And once something is built, I keep it working.
-
Starter sprint
Don't know where to start?
Two weeks: two working sessions to map out what your team does, what should be automated, how it should be automated, and what would be the best way to implement AI into your team.
Rolls into any plan. The sprint is the first two weeks of whichever plan you pick next, and it's credited in full against it.
For people who don't know where to start, or what AI could actually do for them.
Start with a sprint -
A specific automation
For teams who know what they want to automate.
One specific process, automated start to finish, running on your tools: client reporting, lead follow-up, onboarding, whatever eats the hours. The people who own that process are set up and trained to run it.
Automate a specific process -
An AI system that runs your company
For teams that want to get the most out of AI.
A shared AI system inside your team, on your data and your tools, where every person uses AI, people collaborate in it, and their AIs work together. Everyone set up one at a time, trained and supported, with a working session every week.
Training on its own, for teams that already have the tools: sessions for each team, one-on-one setup help and written best practices.
I want the whole system -
Keep it working
I built it. I keep it running.
Two plans, for after any build:
Support planI keep your automations working as your tools and models change, and help with small tweaks. When something comes up, I'm quick to reply.Keep it working and improvingEverything in the support plan, plus I keep building: a regular call, a roadmap, and something new always being added.For when AI is part of how the team runs and you want someone looking after it.
Ask about support
What happens next
No slides, no pitch. You tell me what you want off your team's plate and I tell you honestly whether I'm the right person for it.
- A 45-minute callBring one real piece of work: who does it, how often, and what makes it painful. A short form first, so I've read your answers before we talk.
- A one-page scope within two working daysWhat I'd build, who gets set up, how long it takes, and the price. Yes or no, in writing, nothing to decode.
- We get to workWe pick a start date and go from there: first session with the people who do the work, then one process at a time.
Questions people ask on the first call
How much does it cost?
It depends. I'd love to give you a specific number here, but implementing AI, and what people want from it, is different for every team and every situation. AI work is different for every team: different tools, different processes, and a lot of small one-off details. After we meet, I'll send you a customized pricing sheet with the different options that make the most sense for your team and what you want to do with AI.
How long does it take?
One process, one team, a few weeks. The starter sprint is two weeks. A whole-team build is usually two to three months, with something working inside the first month.
Which AI do you use?
Whatever fits the job and whatever you already pay for: ChatGPT, Claude, Gemini or Copilot. I'm not tied to any one of them, and I won't sell you a new platform.
Do you need access to our data?
Only to the systems the process touches, and only for as long as the build takes. Everything runs inside your accounts, on your tools. You keep every login, and nothing of yours lives on my machine when we're done.
What if it breaks?
Anything I build is covered for the first 30 days. After that the support plan keeps it running as tools and models change, and I'm quick to reply when something comes up.
What if my team won't use it?
That's the part I'm best at. I set people up one at a time and stay next to them until it's a habit. I spend as much time on the people side as the technical side, because that is where most AI projects quietly stop working.
We've been burned by a contractor before. Why is this different?
You deal with me and only me, from the first call to the handover. Nothing is billed by the hour, the scope is one page, and the goal is that you don't need me afterwards.
What happens after the call?
Within two working days you get a one-page scope with the price. If you say yes, we pick a start date and get to work. If I'm not the right person, I'll tell you on the call.
About me
I'm Thomas. I build software, and I spend my time helping teams get real work out of AI.
What makes me unique is that I deeply understand both sides of implementing AI in a team: the technical side and the people side.
The technical side: since the day ChatGPT came out, I've been building with AI every day. Before that, I spent my career in startups and venture capital, founded my own companies, and automated processes in every role I've ever had.
The people side: for five years I've coached people one-on-one, sitting with someone who is stuck and helping them get moving again. I've worked with people through a difficult past, relationship problems, work problems, and everything in between.
I'm hands-on. I don't leave you with a slide deck. I sit with the people doing the work, write the process down with them, build the thing, and stay until they can run it without me.
- Based in
- San Francisco. Works remotely with teams anywhere.