Before You Post That $90,000 Job: A Fractional AI Team for $5,000 a Month Instead

Living page. Updated as the offer and the proof behind it grow. Last updated 16 June 2026.

About the author. Geordie Wardman runs TestVentures.net, fractional AI ops for small companies and enterprise finance teams. Twenty years operating his own businesses, building daily in public since 2026. Bermudian by background, Swiss by residence. Reach: geordie@testventures.net · LinkedIn.

Every few weeks someone I know shows me a job description they are about to post. Operations manager. Or "AI lead." Or the one that gives it away, "generalist, must wear many hats." Ninety thousand dollars, give or take, once you add the social charges and the recruiter fee. They show it to me half proud, half braced.

I always ask the same thing first. What do you actually need this person to do all day?

If it has to do with helping to move a company initiative, get things off the ground, clean up a huge mess or anything vaguely operational, the answer is almost never one job. It is a stack of tasks: clean up the reporting, chase the numbers, wire up some automation, get some AI into the operation, keep the back office moving. A bundle. And a bundle of tasks does not need a full-time body in a seat, or if it does, it's usually such a wide variety of skill sets that you'd need to hire 3 for different roles but not full time.

What you really need, and you may not even know this exists, are two people who are fast, senior, and do not need managing.

So here is the offer this page exists to make. Two of us, two skilled operators, for $5,000 a month. No contract. Cancel any time. You were about to spend $80,000 to $100,000 a year on one employee. USD, GBP, EUR whatever you were thinking. Get two of us for $5,000 a month and run it as long as it is working. A fractional AI team instead of hiring, in plain terms.

What the heck does that even mean?

What is a fractional AI team?

A fractional AI team is two senior operators on a monthly retainer who do the bundle of tasks a small company keeps trying to cram into one job ad.

It is not an agency, where you pay for account managers and interchanging people that come and go from the project. That's so 2023 (and in the age of AI that's decades old). It is not a single freelancer, who goes quiet the week you need them. It is not a SaaS tool you still have to run yourself, or doesn't quite fit that use case you need making it useless. It is two people, working part-time across your operation, building the systems and doing the work, with AI doing the repetitive volume underneath.

The simplest way to think about this: you rent the output, but don't own the employee headcount. In my case the two are me and my CTO-level partner. One of us is closer to the operations and the AI build, the other closer to the engineering. Between us we cover most of what a small firm, or enterprise department tries to hire a "does everything" generalist to do, and we cover it faster because there are two of us and because we have built the same kind of thing before. Countless times.

The work that fits is the rules-based, repeatable half of an operation: reporting, reconciliation, data clean-up, back-office automation, getting real AI into a workflow that is still being done by hand. The half that needs a human judgement call stays with you. We do the volume; you keep the decisions.

The maths: two operators for $5,000 vs one hire for $80,000 to $100,000

This is the comparison every owner is actually running in their head, so let me throw this out on to the table because this is the way I think as well. Ops guy for 20 years, thinking like an owner on every decision that comes across my desk.

A full-time mid-level operations or "AI" hire in a Western market costs somewhere between $80,000 and $100,000 a year all in, once you add social charges, benefits, and the amortised recruiter fee. If you're looking for a virtual controller, CTO or other senior position, more. Then add four to nine months before they are useful at full capacity. Then add the cost of your own time hiring, onboarding, and managing them. Then add the risk that you got the wrong person and have to start again.

A fractional AI team is $5,000 a month, two operators, no contract. Sixty to eighty percent cheaper than the loaded cost of one hire, for two people instead of one, with no ramp and no exit cost.

The honest version: a permanent hire is a known quantity and a fractional AI team is the newer option. But on cost, on speed, and on risk, the maths favours trying two of us first, precisely because trying costs you nothing past the month you are in.

Why it starts in weeks, not quarters

A new hire is a calendar event before it is a contributor. Write the job ad, post it, sift CVs, run interviews, wait out a notice period, onboard. Four to nine months before that person is doing useful work at full capacity is normal, not pessimistic.

How do I know this? Because I was involved as the main hiring person for a team that grew from 4 to 40 in 12 months. I was giving interviews daily, and it's brutal work.

Two of us start inside 10 to 30 days. There is no recruiting, no notice period, no ramp to senior, because we arrive senior. The first month is us learning your operation and standing up the first one or two systems that take pain off your plate immediately. By the time a new hire would still be reading the wiki, the fractional AI team has already shipped something.

That speed is the quiet reason the cost comparison understates the case. The hire costs more and it also costs you the months you wait.

What two operators actually do

The work that fits a fractional AI team is the repeatable, rules-based half of running a small company. Reporting and board packs. Reconciliation and the back-office numbers. Data clean-up. Automating the things still done by hand. Getting real, careful AI into a workflow instead of a chatbot bolted onto the side.

Two concrete examples from work I am building in public right now. For an enterprise finance client I have been building an AI financial controller, the system that does the reconciliation, the accounts-payable capture, and the month-end close prep, with a human signing off the calls that matter. It caught half a bank statement a human had missed in its first real month. Separately, with a finance leader who had never opened an AI tool, we turned a forty-page board pack into one page in under an hour. Same kind of work, smaller scale, fits a small company cleanly.

If what you actually need is someone senior to own your AI strategy and direction rather than do the hands-on building, that is a different role, a fractional chief AI officer, and I have written about that separately. The fractional AI team is hands on the work. The fractional officer is the head above it. Most small companies reaching for a $90,000 generalist need the hands.

Who this is for, and who it is not for

I would rather lose the sale than land the wrong one, so let me be plain about the fit.

This is for the owner or domain expert at a small company who is about to hire because a pile of operational work has built up and they do not know there is another way to get it done. Operations. Automation. Finance ops and reporting. "AI, somehow." The "wears many hats" role that is really five small jobs wearing one job title. If that is the ad you are about to post, two of us will do that bundle faster and cheaper than the hire, and you can stop any month you like.

It is not for everyone. If you genuinely want a person in the building, full-time, who owns one thing deeply, learns your culture, and grows with the company over years, hire that person. If the role needs to hold sensitive responsibility no outside team should hold, hire for it. I will tell you that on the call rather than talk you into a retainer that does not fit. The model works when the job is a bundle of tasks that does not need a permanent seat, and it does not work when the job is genuinely a seat.

It may even be that you bring on an AI fractional team, and that permanent hire at the same time.

What "no contract, cancel any time" really means

No contract means no contract. There is no minimum term, no notice period, no severance. If a month does not deliver, you stop, and we part on good terms. But if you realise 3 weeks later that you actually did have some more tasks for AI specialists, we start again.

The one thing that keeps "cancel any time" honest on both sides is a named scope. At the start we agree the set of tasks the $5,000 covers, written down. That protects you, because you know exactly what you are paying for and can hold it against the result. It protects us, because "two operators for $5,000" cannot quietly become "any task, any hour, forever." When the work grows past the agreed set, we talk about it and you decide. Clear scope is what lets the no-contract promise be real instead of a trap for one side.

This is all managed by a philosophy on a Trello card, one task at a time, or nothing gets done type of mentality.

The proof

A page that says "two people for $5,000" has to answer the obvious doubt: if it were that good, why is it that cheap? The answer is that the price is real and the proof is public.

I publish this business in the open, every day. Revenue, client count, what shipped, what broke. Real numbers, updated as they move, not a testimonial wall. The AI financial controller I mentioned is a real build for a real enterprise client, with real money moving through a real ledger, and I have written up the wins and the misses as they happened. The way I keep a client's data out of a model is written up too. If you want to see how two senior operators actually work before you spend a cent, read the blog this page sits on. The whole point of building in public is that you do not have to take my word for any of it.

One more thing, because trust with other people's numbers is not abstract for me. I also run a fund in my son's name. Handling money and information carefully, for people who are counting on me to get it right, is something I take personally, not as a service-level line.

What clients say

Two decades of running operations for my own companies and other people's. A few of the people I have worked with:

How to start

Book a twenty-minute call. No pitch, no deck. You tell me the job you were about to post, I tell you whether a fractional AI team fits it or whether you should just hire, and if it fits we agree the first month's scope on the spot.

Frequently asked questions

Two people for $5,000 a month, how is that not too cheap to be real?

Because it is two part-time senior operators on a retainer, not two full-time salaries, and because AI does the repetitive volume underneath us so the same two people cover more ground. The price is real. The proof that the work is real is the build-in-public blog this page sits on, where the wins and the misses are both on the record.

Is there really no contract?

Yes. No minimum term, no notice period, no severance. You pay monthly and stop whenever it stops earning its keep. The only thing we write down is the scope of tasks the month covers, which protects both sides.

What if I actually need a full-time employee?

Then hire one, and I will say so on the call. A fractional AI team fits a bundle of tasks that does not need a permanent seat. It does not fit a role where you genuinely want one person, in the building, owning one thing for years. Knowing which one you have is the first thing we work out together.

How fast can you start?

How does next week sound? Really, 30 days or sooner on the first system. There is no recruiting and no ramp, so the first month is learning your operation and shipping the one or two things that save you the most hours first.

What kinds of tasks fit best?

Reporting and board packs, reconciliation and back-office numbers, data clean-up, automating work still done by hand, and getting careful AI into an existing workflow. The repeatable half of running the company. The judgement calls stay with you.

Is my data safe?

Yes, handled properly, and I have written up exactly how I keep a client's information out of a model. Careful handling of other people's data is a first-order part of the work, not an afterthought.

How is this different from an agency or a single freelancer?

An agency sells you account managers and overhead. A single freelancer is one point of failure who goes quiet the week you need them. A fractional AI team is two senior people who do the work directly, build systems that stay with you, and cost less than either the agency retainer or the full-time hire.

Who are the two people?

Me and my CTO-level partner. One of us closer to the operations and the AI build, the other closer to the engineering. The same two people who have built this kind of thing for an enterprise client and are documenting it in public.

If you were about to post that $90,000 job, talk to me first. Book the call, or email geordie@testventures.net. Twenty minutes. No pitch.