Services
Every project runs the same way, in five steps. You see the price before each step starts, and nothing gets built without a signed scope and a deposit. Three decisions along the way, all yours: sign the scope, choose what happens after the MVP, sign acceptance. The first thing I build answers the question you're least sure about.
01Discover
Free. A 30-minute call, then a few follow-up questions.
We start with how the work happens today, step by step: what works, what doesn't, who touches it, what goes in and what comes out. Then the numbers: how often the problem happens, how long it takes, whose time it eats, what it costs when it goes wrong, and what it's worth fixed. And one question that shapes everything after: what are you least sure of? Whether it can be done, whether your people will use it, or whether it's worth it.
If the numbers don't justify a build, I'll say so here, before you've spent anything.
Done when: every question has an answer, or an owner and a date.
02Scope
One written document, with the price. Free when you can already point at the problem.
You get a statement of work: the problem, what success looks like and how we'll measure it, who uses the tool and how, what goes in and what comes out, what's in and out of scope, the MVP in detail and the refinement in outline, the timeline, the price, and what each of us is responsible for. Every requirement traces back to something you told me in discovery, and every deliverable gets a testable definition of done, written like this: you can [do the thing] with [this input] and get [this output] that meets [this measure].
It also includes a risk review. For each step in the workflow: what could fail, what it would cost, how we'd notice, and what we'd do about it. You approve each fix or accept the risk in writing.
The MVP is priced as a range, and you pick the budget within it. More time and money buys a more certain answer.
Who owns the code depends on what we're building. If it's a tool built for your business alone, I expect you to own it. If you're licensing a product of mine, like RQCX, I own the product and you own your data, your instance, and your knowledge base. Either way, the scope says which, in writing, before you sign.
Done when: you've signed and the deposit is in. No build starts before that.
When it isn't clear what the problem is
Sometimes the goal is stuck and it isn't obvious what's in the way. Finding out is real work, so that version of this step is paid. It's called the Systems Roadmap.
What it takes from you: A short questionnaire and a pre-call up front, then five hours on site with you and the people who actually run the work. If you run EOS, have a strategic plan, or set quarterly priorities, I start from those.
What you get: A written roadmap. Your core workflows mapped as they actually run, each problem priced in hours and dollars with the evidence behind every number, and a straight recommendation: what to build first, or that there's nothing worth building yet. If a build makes sense, the roadmap includes the statement of work for it, with the MVP priced as a range.
Cost: $2,500 introductory rate ($3,500 standard). Credits in full toward the build within 45 days.
The document is yours to keep. Every number in it is tagged by how it was verified: counted from your real data, checked on site, or your estimate, marked as such. Nothing gets priced off an unchecked guess.
03MVP
A working version in your hands, fast.
An MVP (minimum viable product) is about 80% of the result for about 20% of the work: real inputs, working output, rough edges allowed. Its job is to answer the question you were least sure about, as fast as possible, so you're deciding on something real instead of a proposal.
Then we decide together: refine it, change direction, or stop. If the MVP shows the idea doesn't hold up, you found that out at the smallest possible price.
Refinement is priced from what the MVP showed, not from a guess, and you approve it in writing before it starts. The MVP is built so refining it never means starting over.
Done when: your uncertainty is answered and you've chosen what's next.
04Refine
The last 20%, built to be used every day.
This is where the rough edges go: the last 20% of the result, which takes most of the work. I build it with the people who'll use it, and they see working versions as it comes together. Scope changes go through a written change process, so nothing drifts and nothing surprises you on an invoice.
Then acceptance: we test with your real data and your real users against the definition of done in the scope. Defects get fixed. You sign off, and the balance is due.
Done when: every deliverable passes its definition of done, and you've signed acceptance.
05Handoff
Yours to run.
You get documentation: what it does, how to use it, how it runs, and what to do if it breaks. I confirm you have access to your data and your instance. Training your people is on you, and the docs are written so you can. A support window follows: defects in delivered work are fixed free for a set period after acceptance. New features, changes to outside services, user error, and bad source data are quoted.
If you'd rather I keep running it after handoff, we can set that up separately.
Done when: you can run it without me.
Frequently asked
- Will this replace my employees?
- It's designed not to. I build tools that take the draining work off your people's plates so they can do higher-value work, and I put that commitment in writing.
- Do customers know they're talking to AI?
- Always.
- What does it cost?
- Discovery is free, and so is the scope when you can already point at the problem. When you can't, the Systems Roadmap is a fixed fee that credits toward the build. The scope comes with the MVP priced as a range, and you pick the budget within it. Refinement is priced after the MVP, from what it showed. A deposit to start, the balance at acceptance. No surprise invoices.
- Who owns what gets built?
- It depends on what we're building, and the scope says which before you sign. A tool built for your business alone, I expect you to own. A product of mine, like RQCX, you license. Your data, your instance, and your approved knowledge base are yours in every case, and they leave with you if we part ways.
- How do you measure the return?
- Discovery puts a number on the problem: hours a week, whose time, what it costs when it goes wrong, what it's worth fixed. The scope sets the success measure from those numbers, and acceptance is tested against it with your real data and your real users.
- What do you need from me?
- Access to the systems the work touches, your real data, time from the people who do the work, and decisions when I need them. Projects move at the speed of those four.
- What if the MVP shows it won't work?
- Then you learned that for the price of the MVP instead of the price of a full build. We decide together: refine, change direction, or stop.
- What happens after it's built?
- You get documentation and a support window for defects. The goal of handoff is that you can run it without me. If you'd rather I keep running it, we can set that up separately.