Tenhaw vs Hiring in-house

Both, in sequence. Hire permanently, which is the right end state and cheaper over any multi-year horizon. Use an interim Embedded Agentic Lead if the board's timeline is shorter than a six-to-nine month search, or if you cannot yet write the job specification accurately. The interim's job includes writing that specification and recruiting against it.

Six to nine months from opening the role to the person starting, then a further three to six months before they can move anything meaningful, because credibility inside a large organisation has to be earned before authority is real. Planning on under twelve months to impact is optimistic.

Currently £180,000–£350,000 base plus equity for a credible candidate, with recruitment fees typically 25–30% of first-year salary on top. That range is our own read rather than a published benchmark. Tenhaw supplies the seat inside an Agentic Design Team at £35,000–£55,000 a month or an Agentic Build Team at £70,000–£85,000 a month, which is more expensive monthly and is intended to run for months rather than indefinitely.

Yes. It is a stated deliverable of the Embedded Agentic Lead and full programme engagements. Recruitment happens while the work is live so that incoming permanent staff join a functioning capability and are onboarded by the person who built it.

It is the common failure, and it usually traces back to the role being given accountability without matching decision rights. The operating model work maps accountability explicitly before the hire is made, which is the single highest-leverage thing you can do to make the role survivable.

Six get discussed and most first workflows need two. An AI engineer builds the system around the model: retrieval, tool calling, the evaluation harness, guardrails, cost and latency. An ML engineer trains, fine-tunes and serves models, which is a different discipline and frequently not what an agentic workflow requires. An MLOps and platform engineer owns deployment, versioning, monitoring and the model upgrade treadmill, and is both the role left out most often and the reason systems stall before production most often. A data scientist frames the problem and builds the ground-truth set that decides what good means, which is the artefact almost nobody has and the only one that cannot be bought. A prompt engineer is the role being absorbed fastest into the AI engineer's job, so hiring it as a standing post is usually a mistake even though the work is real. And an AI product manager owns which decisions move to agents, where a human stays in the loop, and what the workflow is actually for. If you are hiring one person first, hire the product manager or the AI engineer depending on whether your open question is what to build or how. Hire the platform engineer before you have three agents rather than after.

They sit at different points of the same pipeline and merging them into one advert is the commonest reason an agentic hire fails in month four. An ML engineer builds and serves models: training, fine-tuning, feature pipelines, inference performance. An AI engineer builds systems that use models somebody else trained: retrieval, tool calling, orchestration, evaluation, guardrails, cost and latency budgets. A data scientist decides what the problem is and what a correct answer looks like, and owns the ground-truth set that everything else is measured against. Most enterprise agentic work in 2026 is AI engineering with a data scientist beside it, not machine learning, because the models are bought rather than trained. If your job specification asks for all three, you will interview candidates who each meet a third of it, and the one you hire will spend a year discovering which third the role actually needed.

Almost certainly not as a standing role, and the work itself is real. Writing, versioning and evaluating prompts is a discipline with a real effect on output, and it is being absorbed into the AI engineer's job rather than surviving as a separate post, in the same way that nobody hires a dedicated SQL writer. Where it does need naming is in change control, not on an org chart. A prompt is a versioned artefact that changes system behaviour, so a prompt change should trigger the same regression run and the same review as a model upgrade. Treat it as an engineering practice with an owner, not as a headcount line.

Write it from inside the work rather than from a template. The sequence that holds is to start the work, find out which decisions the role actually needs to own in your organisation, and advertise only then. Specifications written the other way round get hired against, and in month four the role turns out to need different authority than it was given, which is the most common cause of early departure in these hires. The other tell is a specification describing three roles at once, merging AI engineering, ML engineering and data science into a single advert. An Embedded Agentic Lead writes yours from inside the work, and recruiting against it is a deliverable of the engagement.

Over three years, no. On that horizon a permanent hire is materially cheaper, the institutional knowledge stays inside the business, and you carry no supplier dependency or commercial conflict of interest. £35,000 to £85,000 a month, depending on whether the seat sits inside an Agentic Design Team or an Agentic Build Team, is poor value as a standing arrangement. It is usually the cheaper path over the nine months of bridging to a permanent hire, because the work starts in weeks rather than after a search and a ramp-up, and the permanent lead then inherits a functioning capability with governance already signed off rather than a blank page.

Often yes. A credible internal candidate ready to step up is one of the clearest reasons not to buy anything from us, because they already hold the domain knowledge and the relationships an external hire spends three to six months earning, and over any multi-year horizon a permanent leader is materially cheaper than we are. What sinks these promotions is accountability without matching decision rights, which is the most common cause of early departure in these roles, so map what the job actually owns before you announce it rather than discovering it in month four. If that person wants support, the useful shape is beside them rather than instead of them.

Frequently your CTO can, and that is the cheaper answer wherever it holds. If the scope is narrow enough for one person to own without organisational redesign, a separate post adds a hand-off rather than capacity. It stops holding when the decisions stop being technical. Which decisions move to agents, where a human stays in the loop and what the workflow is actually for are product and operating-model calls, and a CTO already carrying the estate rarely has the hours for them alongside the day job. Before you create the role, ask whether anything outside engineering has changed the way it works this year.

Some of them, honestly. A candidate who wants to define the function from a blank page will not want to inherit somebody else's design, and if that is who your shortlist is full of, wait for them and give them the blank page. Most senior candidates want the opposite, which is what the engagement is built to leave behind: a workflow already in production, governance signed off, and a job specification written from inside the work rather than from a template. It also makes the interview honest, because you can tell a candidate which decisions the role owns instead of the two of you guessing together.

Only the ones you grant, and they are settled at kickoff rather than negotiated in month three. Month one is spent embedding with real decision rights, because most of this work is deciding what changes: which decisions move to agents, where a human stays in the loop, and what the workflow is for. Be clear about the limit, though. Full-time focus and unambiguous internal legitimacy belong to a permanent hire, and no external lead fully gets them. What an outsider finds easier is the unpopular recommendation, because the exit date is agreed at the start and there is no internal career to protect.

Probably not, if you have a credible candidate in final stages. The bridge exists because a search runs six to nine months on our read of the market, so where yours is close to landing, the honest advice is to keep the money and let them start. It changes if the search has stalled because the specification keeps changing, or if you are close to hiring the wrong shape of person simply to end the vacancy. What you would be buying instead is a specification written from inside the work and a first workflow in production, from the Agentic Design Team rung at £35,000 to £55,000 a month.

Because the candidate pool is thin and expensive, and production experience is rarer than the market makes it sound. Plenty of people can talk credibly about agents having run pilots. Far fewer have put an agentic workflow in front of real users and kept it there, and those who have are rarely looking, which is a large part of why a credible Chief AI Officer commands £180,000 to £350,000 base plus equity on our read of the market. Our own agentic delivery in a regulated estate is a proof of concept being productionised now, and we hold ourselves to the same test by labelling it that way.

Ask what they would put into production first and why, then ask which two roles they would hire behind it. A strong answer separates the disciplines rather than merging them, because an AI engineer who builds systems around bought models is a different hire from an ML engineer who trains and serves them. Ask which of their own examples reached real users rather than a pilot, and who uses them today. Then ask what they changed outside engineering in their last role, because authority rather than technical skill is what these posts most often lack, and it is what makes them survivable.