An independent, technical, skeptical pass over your AI plan — before the SOW gets signed. Most enterprise AI projects fail for the same handful of unsexy reasons. I pressure-test for them up front, so the next $200k comes with a go/no-go memo, not a wish.
$ ./review --slots --quarter
> q3 2026 2 of 2 open
> investment $25,000–$50,000 flat
> duration two weeks
> output go/no-go memo, pre-SOW
> independent. non-promotional. paid only by you._
A vendor or consultancy has put a $200k+ AI implementation in front of you. The deck is sharp. The references check out. The timeline is aggressive but credible. Your team can't find anything obviously wrong with it.
Nobody on the buy-side is paid to find what's wrong before you sign. Vendors are paid to sell. Consultancies are paid to deliver. IT is already underwater. Your own people are too invested in the win to be the ones who say "wait."
So the SOW gets signed, the money moves, and twelve months later the retrospective uncovers the same handful of unsexy failure modes that were visible on page eight of the original proposal — if anyone had been looking.
You don't need another vendor. You need one independent, technical pass over the plan whose only job is finding what breaks.
The review is bounded by design. You bring the proposal. I bring the skepticism. The deliverable is a go/no-go memo you can defend to your board — or hand back to the vendor with the questions they haven't answered yet.
You describe the project, the budget, and the decision deadline. I tell you whether the review is the right fit, what I'll need, and what the memo will and will not cover.
Independent pressure-test of the proposal against the failure patterns that actually sink these engagements. SOW, architecture, data assumptions, vendor claims, team readiness, success criteria. Interviews with the people who'll own the outcome.
A short written memo: the failure modes I found, the ones I didn't, and the conditions under which the plan is buyable. Defensible to a board. Sharable with the vendor. Yours to keep either way.
I've been shipping software since 1998. I've watched cloud, mobile, big data, and now AI go through the same hype curve. The technology changes every few years. The failure patterns don't.
I built the AI Investment Review because the riskiest moment in any AI program is the one nobody is paid to scrutinize: the week before the SOW gets signed. Most of the postmortems I've read could have been the pre-mortem.
I sell no software. I take no referral fees. I'm only paid by you.
$ whoami
Scott Pierce
$ cat experience.txt
25+ years shipping software
10+ years in production AI
4 technology hype cycles survived
$ cat allegiance.txt
"On your side. Not the vendor's.
Not the consultancy's. Yours."
$_
If you've already signed the SOW — the review is a pre-commitment tool. Mid-implementation rescues are a different engagement and usually a different person.
If you want validation, not pressure — the memo will name what's wrong. If the answer you want is "this looks great," you don't need me.
If there isn't $200k+ on the table — the cost-to-decision ratio doesn't work below that threshold. Smaller commitments deserve different tools (see below).
If you need the answer this week — two weeks is the minimum honest review. Anything faster is a glance, and a glance is what got the proposal this far.
You sign with eyes open — the failure modes are named, mitigated, and tracked. The board update writes itself.
You don't sign — the memo names the specific conditions the vendor would have to meet. You keep the $200k+ and the leverage.
You renegotiate the SOW — the most common outcome. Same vendor, sharper scope, smaller blast radius if it does fail.
You become the leader who pressure-tests — once, then by default. The org around you starts asking the same questions on every commitment that follows.
The $200k+ that's already gone — and the second $200k that gets approved to "save the first."
The board update full of euphemisms — "foundational work," "longer than expected," "pivoting to phase two."
The competitor announcement — the one you'll read 18 months from now that proves it was doable, just not by this vendor with these assumptions.
The credibility tax — the next AI proposal you bring to the board lands in a colder room than the last one.
The budget is the obvious loss. It's not the biggest one.
30-minute intake call. No fee, no pitch, no pressure. We decide together whether the review is the right fit before anyone signs anything.
Q3 2026: 2 of 2 open.
A 5-minute self-check on whether your team can describe what they're building clearly enough to build it. The methodology that shows up inside the review.
The failure patterns I keep finding, in long form. For the people who want to recognize them before they show up in their own SOW.
Short-form thinking, ongoing. Best place to stay current between engagements.