Your AI project stalled. The license is still renewing.
Three to four weeks. One honest assessment. A path forward you can actually take.
Something got built. A vendor's professional services arm shipped it, or another consultant did, or your internal team prototyped it during a quiet quarter. It went live. People used it for a while.
Then the energy left the room. The team that championed it moved on, the workflow it was supposed to fix changed, or the outputs stopped being trustworthy and were quietly worked around. Adoption dropped. The dashboard meant to prove ROI hasn't been opened in 3 months.
Meanwhile, the contract auto-renews. Killing the project means admitting the original decision was wrong, so it doesn't get killed. It just sits there, costing money and consuming a small amount of attention every week, defended by no one and removed by no one.
This is the most common AI project state in established businesses right now. More common than success. More common than visible failure.
AI Pilot Recovery is a three-to-four-week assessment of one stalled or underperforming AI project. The deliverable is a written recommendation memo that tells you whether to fix it, scope it down, or shut it off — plus a 90-day path forward you can hand to whoever owns the work.
The memo is built around three options, in this order: fix the existing system so it actually delivers what it was meant to; scope it down to the part that's working, and shut off the rest; or decommission it entirely and move on, with a clear migration plan for whatever it touches.
I don't have a preference among the three. The recommendation goes wherever the evidence points. Sometimes that's a clean fix. Sometimes the right answer is the one nobody wants to put in writing — which is why the write-up matters.
This engagement is for owners or operators in one of three situations.
You inherited an AI project from a predecessor. You don't know whether it's working. You don't know whether it's worth the contract you keep signing. You need someone with no political stake in the original decision to look at it honestly.
You championed an AI project that hasn't delivered, and you need a credible outside view before you decide what to do. Killing your own project is hard. An external assessment makes it easier to act on the answer, no matter how it lands.
You have a vendor or consultant relationship that started with a build and has drifted into a low-grade ongoing engagement nobody can quite define. You want to know whether the relationship should continue, get rescoped, or end.
If your situation looks like any of these, this engagement is built for you.
Three to four weeks. Four deliverables.
Discovery and Evidence Gathering
We start with a working session covering the project's history. What was it built to do? What's actually happening now. Who uses it, who's stopped using it, who never used it. What does the current contract obligate you to? Then I gather evidence directly — usage data, sample outputs reviewed for quality, and interviews with two to four people across the team.
Assessment
I evaluate the project against three questions. Is it doing the job it was built to do, to the required quality? If not, can it be fixed at a reasonable cost and within a reasonable timeline? And is the underlying need still real, or has the business moved on? Each question gets a documented answer with supporting evidence.
Memo Drafting and Path Forward
I write the recovery memo. The recommendation up front. The reasoning and evidence in the body. A 90-day plan for whichever path is recommended — fix, scope down, or shut off — with named actions, named owners, and named decision points.
Walkthrough and Handoff
A one-hour walkthrough call, then a handoff document for whoever owns execution. Most engagements close in three weeks. Four-week engagements happen when the project touches multiple systems or when contract exit requires coordination across legal and procurement.
What you leave with.
The Recovery Memo.
Ten to twenty pages, depending on project complexity. Recommendation up front, evidence and reasoning in the body, and risks and assumptions explicitly named.
The 90-day path forward.
One document with named actions, owners, and decision points for the recommended path. Written so it can be handed to your operations lead without needing further translation.
The contract review.
If the recommendation involves changing or ending a vendor or consulting relationship, you get a one-page summary of the contract terms relevant to that change, common exit gotchas, and language to use in the conversation.
The walkthrough.
One hour to talk through the memo, push back on the reasoning, and adjust if anything in the writeup needs sharpening before you act.
This isn't a sales mechanism for me to take over the project. The memo's job is to give you the right answer for your business, which is sometimes “kill it and don't replace it.” If the recommendation includes new build work, that's a separate, scoped engagement, and you're free to take the memo to anyone you want for execution.
It isn't a vendor evaluation. If you're trying to choose between alternative tools to replace a stalled one, the Investment Memo is the right engagement format.
It isn't an audit of your full AI footprint. If multiple AI projects are running in your business and you don't know what state any of them are in, the Exposure Audit covers that ground first. The Pilot Recovery engagement assumes you've already identified the one project that needs assessment.
And it isn't a free second opinion. Buyers sometimes want to validate a decision they've already made. That's a fine reason to engage, and the work is the same regardless. The memo will reflect the evidence, which may or may not align with what you hoped to hear.
$5,000–$9,500.
Depending on project complexity, the number of stakeholders to interview, and whether the recommendation involves contract exit work. Every engagement gets a fixed quote before work begins. No hourly billing. No scope drift.
Every engagement begins with a Discovery Call to confirm the project is the right shape for this work. If the situation calls for a different starting point, I'll let you know, and we'll talk about which engagement fits best.
The hardest part is naming the situation honestly.
Book a Discovery Call. Thirty minutes. No pitch.
Book a Discovery Call →