Most “what’s it like to work with us” pages are marketing. This one isn’t. Here’s an honest, month-by-month version of what AI Office actually feels like from a client’s seat.
Why this page exists
Owners thinking about an AI partnership tend to ask the same questions on the second call:
- How often are we actually meeting?
- What do I get vs. what do my team get?
- When will things show up vs. when will I have to wait?
- What does it feel like when something goes wrong?
This is the version of those answers that doesn’t sound like marketing. It’s drawn from how our clients actually describe the rhythm to their peers.
Month 0: Discovery
Before anything is signed. A 30-minute call. We ask about your business, your pain, where AI might fit. You ask about us. If there’s mutual fit, we send a 2-page Discovery Letter the next morning with a recommended tier and a rough first-90-day arc.
What you do: a 30-minute call, a 5-minute read. What we do: 30 minutes on the call, 2 hours synthesizing.
The Discovery Letter is intentionally short and intentionally specific. If it doesn’t tell you what we’d actually do, that’s a problem with our diagnosis, not your decision.
Month 1: Kickoff and roadmap
Week 1: An in-person kickoff at your office (if geography allows). 90 minutes with you and your senior team. We listen more than we talk. We leave with read-only access to your systems and a clear sense of what we’d recommend.
Week 2: Async work. Calls with key people. A workflow scan. A look at 10-30 examples of the work in question.
Week 3: A 1-page roadmap with 3 priorities and a recommendation on which to start. We share it in writing before we walk through it. You push back. We adjust.
End of month 1: The roadmap is agreed. The first build has started.
What you do: ~12 hours of senior team time (CEO, COO, the function leads). What we do: ~50-70 hours across our team.
What it feels like: busy but productive. The roadmap clarifies things you knew but hadn’t named. The team is starting to believe.
Month 2: First build
Weeks 5-7: The engineer is heads-down in the build. You see weekly Looms (5-10 minutes each). You see a 30-minute demo at week 6. Edge cases surface; you flag them; the engineer adjusts.
Week 8: The workflow goes into production for a small user group. Monitoring is live. The senior strategist is at your office for a half-day reviewing how it’s running.
What you do: ~6 hours of senior time. Most of the burden is on the function lead and 1-2 power users. What we do: ~80-100 hours across our team. Heavy on engineering.
What it feels like: real. Something concrete is happening. The skeptics in your team start softening. The champions are emerging.
Month 3: Refinement and production
Weeks 9-12: The workflow is in actual use. Quality is being measured. Refinements happen weekly. The strategist has a 45-minute call with you each Monday. You get a written briefing each Friday.
Day-90 review: A 90-minute meeting. We walk through what was built, what’s working, what’s not, and what’s next. We bring an updated roadmap. You leave with documented ROI on the first workflow.
What you do: ~10 hours total this month. What we do: ~60-80 hours, shifting from build to operations and the start of the next priority.
What it feels like: the first ROI conversation is the inflection point. When the number is real, the relationship shifts from “we’re trying AI” to “we’re doing AI.” The brand promise — that we pay for ourselves — starts becoming visible.
Months 4-6: Second workflow, rhythm sets in
The rhythm: Monday standing call, Friday briefing, monthly review, quarterly bigger conversation. The second workflow ships somewhere in month 5. By the end of month 6 you have 2 workflows in production and 1 in flight.
What you do: 8-12 hours per month is the steady-state senior time investment. What we do: 60-100 hours per month, depending on tier (Operator vs. Embedded) and Sprint activity.
What it feels like: settled. The team has internalized the cadence. AI is a normal part of how the business runs. New opportunity discovery comes from your team as often as from us.
Months 7-12: Tier upgrade or stable, deeper builds
For most clients at Operator who started small, this is when the tier conversation happens. Either the build queue has grown to a level Operator can’t sustain (upgrade to Embedded), or the rhythm is stable and Operator continues to fit (stay there).
Around month 10, the first “deeper” build conversation comes up. Often this is a multi-quarter program — a multi-Sprint sequence on something operationally larger.
What you do: maintain the cadence. 8-12 hours per month most months. What we do: scale to match the work. By month 12, often a 3-person team across strategist, engineer, and a second engineer for deeper builds.
What it feels like: normal. The drama is gone. The work is producing. The conversations are about strategy and prioritization, not crisis management.
What it feels like when something goes wrong
Things go wrong. Here’s what that actually looks like.
A workflow produces a bad output that reaches a customer. Illustratively, here’s how this plays out: we notice within hours (because monitoring is in place), notify the client immediately, help craft the customer response, and have a fix in production quickly. Post-incident, we document what happened, update the system to prevent recurrence, and add the failure mode to our training for new engineers.
A build runs long. This has happened more often than the customer-facing incident. We tell you when we know. We don’t pad the estimate; we explain why and what we’re doing. We don’t bill extra for our own miss.
A client outgrows us or wants to bring it internal. This has happened too. We help with the transition. We share documentation. We don’t sabotage the handoff. We’ve actually been brought back later by some of these clients on lighter tiers because the relationship was healthy at the end.
What it feels like in all three cases: the partnership doesn’t break under stress. The honest conversation happens early. The cleanup is fast.
What it feels like at the end of year 1
Here’s the version we aim for, described illustratively: AI is now just part of how the business runs — not talked about as a project anymore. Several things are in production that the team couldn’t have built themselves in the same timeframe. The team is more capable than it was. The numbers work. And the partnership feels like a relationship, not a vendor agreement.
That’s the version we’re aiming for. We don’t hit it every time. When we don’t, it’s usually because the underlying sponsor commitment wasn’t there, or the problem wasn’t the right one. When we do hit it, the relationship typically continues for years.
A 30-minute conversation
If the rhythm above sounds like what you’d want — and you have a specific business in mind — the 30-minute call is the right next step. We’ll tell you straight whether AI Office fits and what your first 90 days would actually look like.