CPO of One
[ 00 — CPO of One ]

Your roadmap needs proof, not the loudest request.

We prototype your ideas with AI so users try them before engineering commits a sprint, and your PMs learn to do it.

What you get Book a call

[ 01 — The requests ]

Sales wants it. Support wants it. Who tested it?

Roadmaps get set by whoever asked last, or loudest.

[ 02 — The break ]

Then prototypes got cheap.

AI turns a prompt or a spec into something users can click.

[ 03 — The sort ]

Every request becomes a hypothesis.

Interviews. Tickets. Sales notes. Usage data.

[ 04 — The bet ]

One bet. Worth testing.

[ 05 — The product leader ]

You decide what gets built.

AI drafts the options. Users react. The call stays yours.

[ 06 — The versions ]

Build five versions. Keep the one they use.

Put working options in front of users, not slide decks.

[ 07 — Craft and pipeline ]

We sharpen the calls. We build the pipeline.

Teach

Product craft for cheap prototypes: prompting a prototype, writing specs an agent can build from, running evals and synthesizing research, practiced on your own roadmap.

Build

The product pipeline itself: prototype setups, interview tagging, eval suites and release checks, built beside your team so they can run it.

[ 08 — Signals ]

Your product, inside someone else’s chat.

  1. Instead of fixed UI, AIforce empowers anyone to build composable, intelligent, live interfaces, wherever work happens.
    Salesforce AIforce announcement, September 15, 2026 Source ↗
  2. Meta received more than 1,500 applications from developers to build connectors for its Muse agent in less than a week.
    Alexandr Wang, Meta Company-reported, via TechCrunch, September 23, 2026 Source ↗
  3. We’ll maintain the legal configuration so builders can focus on their own products and workflows.
    OpenAI, launching Astra for Law Launch statement quoted by Artificial Lawyer, September 18, 2026 Source ↗
[ 09 — Two kinds of user ]

Design for people. And for their agents.

Structured data, clear permissions, and a way to know who is asking.

[ 10 — Next release ]

The next feature may have no screen.

Voice and avatars are now a configuration choice. Design the trust first.

[ 11 — Hours or coaches ]

Hours for the decision. Coaches for the build.

Consult

$400 per hour

Reserved in 5-hour blocks

Hours with our coaches on a live product call: which bet to test, how to prototype it, and how to read the result.

Reserve hours → Details

On Build, we get projects up and running within a week, depending on complexity and how quickly you give us access to your systems.

[ 12 — Begin ]

Bring us your riskiest idea.

Which way in?

Thanks. Choose a time below, or watch your inbox for our reply.

Pick a time →

[ Questions ]

What is CPO of One?
CPO of One (cpoofone.com) is a site from Utlyze for heads of product and product managers. It covers product discovery and AI-feature decisions inside a company: prototyping with AI, specs agents can act on, evals and user research.
Who is CPO of One for?
Heads of product and product managers at small and mid-size software companies who sit between stakeholders and users, and who now have AI tools that can turn an idea into a working prototype.
Do product managers need to code now?
Not to ship production code. AI prototyping tools let a product manager build a working prototype and test an idea with users before engineering time is spent. Production code still belongs with engineers and their review process.
What is AI prototyping?
Generating an interactive prototype from a prompt or a written spec, then putting it in front of real users to see what they do with it. The prototype is a question, not a release.
How do I write a spec an AI coding agent can build from?
State who it is for, what done looks like, the data and screens involved, and what must not happen. Add examples of good and bad results. Keep the parts a person must decide, such as scope and trade-offs, in plain sentences at the top.
How do I evaluate an AI feature before launch?
Collect real inputs, write down what good and bad outputs look like, run the feature against that set (these are evals), and review the failures with users. Rerun the same set whenever the model or prompt changes.
Can AI do user research?
It can transcribe, tag and summarize interviews, and help you spot patterns across many of them. It cannot replace talking to users, and its summaries should be checked against the recordings.
How should we price AI features?
Start from what each use costs you to run and what the result is worth to the customer, then choose between bundling, usage limits or a separate tier. Test the price with customers the same way you test the feature.
How much are Consult and Build?
Consult is $400 an hour, reserved in 5-hour blocks ($2,000 minimum). Build is $15,000 a month. Each month, pick up to 5 business processes for our build-with-you coaches to work on. They build your AI systems with you and teach your team to run them. If you ever leave, the systems stay with you. Agents work only on your systems, and a shared 24/7 on-call rotation answers, with response times in writing.
How is this different from Business of One?
CPO of One covers product decisions inside a company: what to test, how to prototype, how to judge AI features. Business of One is for owners who run an AI-first business as a whole. Both are run by Utlyze, and both offer Consult and Build on the same terms.

Built on the Of One method: ask the right questions, map what you find, make the best move. of1.ai ↗