Skip to content
ClelandCo

Vol. II — AI Strategic Advisor

Currently booking

A senior second opinion on what to build, what to buy, and what to stop.

Your team can build it. Nobody has asked whether you should.

A standing advisory seat for teams already building with AI: a scored roadmap, honest build-versus-buy calls, vendor evaluation, and review of what you are actually shipping. Your leadership still decides. I make sure the decision is informed.

Discuss an AI Adoption Assessment (opens in a new tab)

Briefing

The teams that need this are not stuck. They are shipping. Engineers are wiring up models, features are going out, and the demos are good. What is missing is anyone senior enough to ask whether the evaluation story holds, whether the vendor about to be signed will survive contact with production traffic, and whether the roadmap is a list of outcomes or a list of features that happen to use a model.

Those questions get asked eventually — usually by a customer, a security review, or an incident. Asking them early costs a fraction as much and is considerably less embarrassing.

The engagement runs as a standing weekly working session with the people doing the build, not a monthly briefing for the people funding it. Real pull requests, real evals, real contracts. Between sessions I am available for the decisions that cannot wait a week. Each month you get a written record of what was decided and what it depends on, which turns out to be the artifact teams reach for most.

The goal is that your team stops relitigating the same three decisions, and can run the criteria themselves after I am gone.

Three ways forward

Three options. One recommendation.

01

Hire a full-time AI lead.

The right answer eventually, and a long search for a role most companies cannot yet scope precisely. Hiring for a job description you are still writing is how you end up with an expensive mismatch.

02

Bring in a consultancy.

You get a thorough document and a bill. The analysis is usually sound; the problem is that the engagement ends at the deliverable, and nobody is around when the plan meets the first sprint that disagrees with it.

03

Add a senior second opinion.

A standing seat with the people building, sized to the judgment gap rather than the ownership gap, ending in criteria your team keeps.

The engagement ladder

Two rungs. One clear path.

Most engagements start with the assessment. It is self-contained and yours to act on regardless of what happens next — including when it recommends doing nothing.

AI Adoption Assessment

The front door

$6,000

One time, 1 to 2 weeks

A written read on where AI is worth spending, what has to be true before anything ships, and what to do first. The same assessment starts both the advisory and fractional tiers; it ends with a straight recommendation about which, if either, is worth it — including when the answer is neither.

AI Strategic Advisor

The engagement

$5,000–$8,000/mo

20–30 hrs/mo · Three-month initial term, reviewed each quarter

An advisory seat for teams already building with AI: a scored roadmap, honest build-versus-buy calls, vendor evaluation, review of what you are shipping, and evaluation criteria your team can run without me.

Discuss an AI Adoption Assessment (opens in a new tab)

AI Adoption Assessment

$6,000

One-time

A written read on where AI is worth spending, and what to do first.

Format

  • Interviews with the people building and the people funding, held separately, because they rarely describe the same roadmap
  • A use-case inventory scored by value and feasibility, with an explicit list of what not to do — including funded pilots that should be killed
  • A readiness and governance gap assessment covering data handling, evaluation, deployment, and monitoring, written so you could hand it to a customer or an auditor
  • Build-versus-buy guidance per use case with the criteria written down, plus a review of the vendor commitments already in flight
  • A prioritized roadmap and a straight recommendation on whether ongoing work is justified, including when the answer is no
Discuss an AI Adoption Assessment (opens in a new tab)

The AI Adoption Assessment, $6,000 one time.

This is right for you if…

You have a team building AI and a roadmap nobody senior has pressure-tested. You are about to commit budget — to a vendor, a hire, or a quarter of engineering time — and you would like someone who has done this to look at it first. It is the same assessment the fractional tier starts with; what differs is the recommendation it ends in.

AI Strategic Advisor

$5,000–$8,000/mo

20–30 hrs/mo

A standing weekly seat with the people doing the build.

Format

  • A use-case inventory scored by value and feasibility, including an explicit list of what not to do and why.
  • A written build-versus-buy recommendation per use case, with the criteria used to reach it so the call can be re-examined later.
  • Vendor evaluation: a shortlist, the questions to ask, and the failure modes to test for before anyone signs.
  • A review of what your team is already building, covering data handling, evaluation, deployment path, cost, and monitoring.
  • Evaluation and governance criteria written so your team can run them after the engagement ends.
  • A rehearsal plan from the AI Contingency Framework — the red-team, failure-injection, and misuse exercises to run before anything ships, designed for your team to execute.
  • Working sessions with the engineers and leads doing the build, so the upskilling lands with the people who need it.
Discuss an AI Adoption Assessment (opens in a new tab)

A standing weekly working session with the people doing the build, async availability between them, and a written monthly update.

This is right for you if…

Your engineers are capable and have never taken a model past a prototype. Decisions keep getting relitigated because nobody wrote down why the last one was made. You want senior judgment in the room every week, not a document once a quarter.

Fit

Worth knowing before you call.

Signals this is your tier

Executives and founders with a team already building AI who need someone senior to pressure-test the plan without taking the wheel.

  • Your engineers are shipping AI features and nobody senior is checking the evaluation story.
  • You are about to sign a vendor contract and cannot tell whether the demo will survive production.
  • The roadmap is a list of features rather than a list of outcomes.
  • Your team is capable but has never taken a model past a prototype.

Where this is the wrong call

  • Teams with nobody to execute. Advice with no implementer behind it turns into a document. Take the fractional seat instead.
  • Organizations that want someone accountable for delivery. Advice is not ownership, and this tier does not pretend to be.
  • Anyone who wants a deck for the board. The output is decisions and written criteria, not slideware.
  • Companies with no AI work under way and no named business problem. Start with the assessment and stop there if it says to.

What happens first

The opening phase, step by step.

Your team stops relitigating the same decisions. There is a written roadmap, a build-versus-buy call on each item with the reasoning attached, and evaluation criteria everyone agreed to before work started.

  • Weeks 1–2

    Inventory and honest scoring.

    Every AI use case under way or proposed, scored by value and feasibility, with an explicit list of what not to do and the reasoning attached. Most teams find at least one item they had already funded and cannot defend.

  • Weeks 3–6

    Decisions, in writing.

    Build-versus-buy on each surviving item, with the criteria recorded so the call can be re-examined when circumstances change. Vendor shortlists, the questions to put to them, and the failure modes to test before anyone signs.

  • Weeks 7–12

    Review and handover of criteria.

    Working sessions move to what your team is actually shipping: data handling, evaluation, deployment path, cost, and monitoring. The evaluation and governance criteria get written down in a form your team can run without me.

Against the alternatives

Three tiers. One ladder.

DimensionAI Visibility LeadAI Strategic AdvisorFractional CAIO
Price$1,500–$2,500/mo$5,000–$8,000/mo$10,000–$15,000/mo
Time per month10–15 hrs/mo20–30 hrs/mo50–70 hrs/mo
Starts withAI Visibility Baseline, $1,000AI Adoption Assessment, $6,000AI Adoption Assessment, $6,000
Who owns the outcomeYou own the outcome. I measure it and close the gaps.You own the outcome. I pressure-test the plan.I own the outcome and report on it.
Decision rightsNone. I recommend and execute inside an agreed scope.Advisory. Your leadership decides.Full, inside an agreed mandate and budget.
Hands on keyboardYes. Profile, schema, and citation work is included.Review only. No implementation in this tier.Yes, on the pieces where a handoff would kill the work.
Best forGetting found and cited correctly by AI search.A capable team that needs senior judgment.An AI function with nobody accountable for it.

The operator

Who you're working with.

I am Jeremy Cleland. Fourteen years in uniform, eleven of them in Special Operations as an 18D Special Forces medic, then a master's in Artificial Intelligence in 2025 and a second career building systems that have to work when someone is depending on them.

Special Operations medicine is a useful background for this work for an unglamorous reason: the job is making decisions with incomplete information, under a deadline, where being confidently wrong is the failure mode that costs the most. That is also the job of running an AI function in 2026.

I build as well as advise. The roadmaps I write are constrained by what I know I would have to implement, which tends to keep them honest about what is actually shippable this quarter.

Most AI advisors are strategists who have never shipped a model, or engineers who have never sat in a leadership meeting. The useful part is doing both.

Read the longer version

What this engagement does not include.

Implementation. I read your code, your evals, and your deployment path and tell you where they break. I do not write the fix. Building is a different commitment — either the fractional seat, where I own delivery, or a separately scoped implementation.

Accountability for delivery. Advisory is not ownership and this tier does not pretend otherwise. If what you need is one person answerable for whether the AI work produces business results, that is the tier above and it is priced accordingly.

Board slideware. The output is decisions and written criteria. I will happily join a board conversation and answer for the reasoning; I will not spend advisory hours making a deck look finished.

Unlimited availability. The retainer is sized in hours per month, not in vague access. That is deliberate: a retainer that promises everything gets rationed quietly, and you find out which parts were fictional at the worst moment.

Questions

Asked and answered.

What does an advisory month actually look like?
$5,000 to $8,000 per month for roughly 20 to 30 hours a month. That buys a standing weekly working session with your team, availability between sessions for the decisions that cannot wait a week, and a written monthly update recording what was decided and what it depends on.
What does the AI Adoption Assessment cost and include?
$6,000, one time, over one to two weeks. A use-case inventory scored by value and feasibility, a readiness and governance gap assessment, build-versus-buy guidance, and a prioritized roadmap. It is self-contained and yours to act on regardless of what happens next.
Do you write code in this tier?
No. I read your code, your evals, and your deployment path, and I will tell you where they break. Building is a different commitment: either the fractional seat, where I own delivery, or a separately scoped implementation. Charging advisory rates for implementation time serves neither of us.
How is this different from the Fractional CAIO tier?
Decision rights. Here I advise and your leadership decides; the accountability stays with you. In the fractional seat I own the roadmap, sit in leadership meetings, and answer for the results. Advisory is the right call when you have a capable team and a gap in senior judgment, not a gap in ownership.
Can you train my team rather than just advise the executives?
That is most of the value. The weekly session is run with the engineers and leads doing the work, not only the people funding it. Reviewing real pull requests, real evals, and real vendor contracts teaches more than a workshop, and it leaves criteria behind that the team can apply on their own.
What if the assessment says we should not be doing this?
Then it says that, in writing, and you have saved considerably more than the assessment cost. A recommendation to stop is a legitimate output. An assessment that always concludes you need a retainer is a sales document, not an assessment.
Do you cover what happens when the system fails?
Yes, and it is a larger part of this tier than most advisory engagements make it. The AI Contingency Framework's planning phases run here, and I design the rehearsal — prompt injection, data poisoning, failure injection, bias evaluation, misuse simulation, and a disaster-recovery walkthrough — for your team to execute against your system.

Bring the decision you are avoiding.

The fastest way to find out whether this is useful is to put the hardest open question in front of me. If the honest answer is that you do not need an advisor, that is a short call and a straight answer.

Discuss an AI Adoption Assessment (opens in a new tab)