Tenhaw vs AI boutiques

A build shop delivers working AI software. A transformation consultancy changes how the organisation operates so that software is actually adopted: redefining roles, moving decision rights, rewriting governance and managing the resistance that follows. Most failed agentic programmes have working technology and an unchanged organisation.

Ask for named clients with checkable outcomes, not anonymised logos. Ask who specifically will be in the room and what they personally delivered. Ask what happened after they left the last three engagements. Ask whether they will publish a price. Ask what they would decline to do. Firms that cannot answer those five questions concretely are usually selling capacity rather than capability.

Because agentic transformation is a change programme with AI in it. The hard parts (moving decision rights, redesigning roles, managing incentive conflict, sustaining adoption past the initial enthusiasm) are the same problems delivery transformation has always faced. A firm that has never made ten thousand people work differently will meet those problems for the first time on your programme.

No. For a narrowly scoped technical build, a specialist shop will usually be cheaper and often better. Tenhaw is priced for organisation-level change where the operating model, the engineering and the adoption all have to move together.

Six come up repeatedly, and they are different shapes rather than versions of the same firm. Faculty, in London, works across defence, national security, health and financial services. Mind Foundry, in Oxford, is a university spinout leading with machine learning for defence and national security. Aiimi, in Milton Keynes, sells its own platform alongside services across enterprise data, search and governance. Advancing Analytics is an elite Databricks partner and a Microsoft advanced specialist. Kortical sells an AI platform and consulting together. Datasparq covers strategy and implementation. Those descriptions come from each firm's own published material. Ask all seven of us the same three questions: who will be in the room, what the first paid engagement ends in, and whose repositories the code lands in.

Who owns the thing at the end, and where the supplier's margin comes from. A firm selling its own platform has a commercial interest in your architecture running on it, which cuts both ways. Part of what you buy keeps improving without you paying for it, and part of your estate is theirs to version. Three of the six UK firms named here sell a platform or product of their own. Tenhaw sells no product, resells no models, platforms or licences, and takes no margin on any of them, so no part of your run cost is revenue for us. Having no compounding asset to amortise is one reason we are not the cheapest per day. Ask any supplier what you could change a year after they leave, without them.

Only if building or tuning models is the actual work. If you need a specific model built or fine-tuned and your operating model is not in question, a research-led firm is the better buy and we will say so on the call; for specialist modelling we partner. Where that is not the requirement, a research bench is rarely what unblocks a programme. The harder half is deciding which decisions move to agents, making those decisions real in working software, and getting people to use them. That is what Tenhaw sells, with the operating model, the engineering and the adoption in one squad of three senior people.

Usually not, because the seam between them is where agentic programmes stall. Strategy boutiques hand over a thesis and a roadmap and depart before implementation. Build shops deliver working software into an organisation whose roles, incentives and governance are unchanged, which is why the software often goes unused six months later. Those three parts hold each other up. The operating model decides which decisions move to agents, the engineering makes them real, and adoption work makes people use it, so Tenhaw keeps all three in one squad. If you do split the work, decide who owns adoption, since the strategy firm has left by then and the build shop rarely treats it as in scope.

It depends on how much of the difficulty is domain knowledge. Where the requirement is deep vertical work, clinical, legal or quantitative, domain specialists lead. Several firms here point that way, with Faculty positioning itself across sectors including defence, national security, health and financial services, and Mind Foundry leading with defence and national security. Buy the specialism when that knowledge would take us weeks to acquire and your operating model is not in question. Where the blocker is that roles, decision rights and adoption all have to move, a record of making thousands of people work differently matters more, which is what a decade of delivery transformation at HSBC, Microsoft, F1 and Anglo American gives us.

Ask each firm on your shortlist directly, because the default answer is no. Technical firms typically treat organisational resistance as out of scope, escalate it to the sponsor and continue building, and the sponsor was usually hoping the supplier would handle it. The gap matters because transformations stall on incentives and territory far more often than on technology. At Tenhaw, managing the politics is an explicit part of the Embedded Agentic Lead role rather than an unfortunate distraction from it, and it draws on experience of managing transformation politics inside a 100,000-person organisation. Whichever firm you choose, ask where organisational resistance sits in their scope, since most technical firms will hand it back to you.

Weigh the record rather than the founding date. Behind Tenhaw sits a decade of landing delivery transformation at HSBC, Microsoft, Sky, F1, Discovery, Greggs, Yondr and Anglo American, twelve engagements written up in full with checkable outcomes, and a Portfolio Management Office built from nothing at Tecknuovo that governed 19 projects including HMRC, the MOD and Thames Water. Current agentic evidence runs to a two-week proof of concept inside a live regulated insurer, being productionised now, and more than twenty agentic products built through Velocity84, our build lab. Agentic transformation is a change programme with AI in it, and the change discipline is the half most technical firms hand back to you.

Only if the hard part is platform engineering on that stack. Advancing Analytics describes itself as an elite Databricks partner and a Microsoft advanced specialist, and where your estate is built on those platforms and the work is deep engineering inside them, that certified depth is worth buying and we will say so on the call. Tenhaw is not a platform partner business, since we sell no product and no licence, resell nothing, and take no margin on models or platforms. No part of your run cost is revenue for us. What we bring instead is the operating model, the engineering and the adoption in one squad.

Sometimes, and if you do, give both firms the same problem, the same access and the same fortnight, then judge on what runs rather than on what was presented. The cost buyers underestimate is not the two fees, it is your own subject-matter experts answering the same questions twice while doing their day jobs, and they are the scarce resource on any agentic programme. The cheaper version is to buy one fixed-price piece of work from your first choice and keep the second warm. Our AI Readiness Audit is £44,000 for four weeks, standalone, ending in working prototypes, with no obligation to continue.

Ask what would happen if the technology arrived tomorrow. If your teams could not build or run it, the blocker is capability, and the buy is whichever firm holds the depth your stack actually needs, in model engineering or in platform engineering. If something already works in a pilot and nobody beyond the team that built it uses it, the blocker is adoption, and more engineering will not shift it, which is where a decade of delivery transformation at HSBC, Microsoft, F1 and Anglo American earns its keep. Name the people whose job changes on the Monday after go-live, and name who has the authority to decide that it changes. If nobody can, capability was never the problem.

Often yes for the first working system, which is a real advantage and worth weighing. Three of the six UK firms named on this page sell a platform or a product of their own, and another publishes its own accelerators, so a workflow shaped the way their platform expects can be stood up quickly, and part of what you buy keeps improving without you paying for it. The trade is that part of your architecture is then theirs, and the run cost carries a licence. Tenhaw sells nothing of the kind, so what we build runs in your infrastructure under your policies, and the code and documentation are yours.

Real decision rights, which is the part most firms are never given. In practice that means a sponsor who spends time on it, the authority to redefine roles, move decision rights and rewrite governance, and access to the people whose day-to-day work changes. Without that you get a build shop outcome whatever you paid. Working software lands in an organisation whose roles, incentives and governance are unchanged, which is why so much of it sits unused six months later. Grant it and the work speeds up, because decisions get settled in the room rather than escalated. If you cannot grant it yet, that is far better known before you shortlist than in month three.

Not on trust, so check it. Everything this page says about Faculty, Mind Foundry, Aiimi, Advancing Analytics, Kortical and Datasparq comes from each firm's own published material, read in July 2026, with a link to the page it came from, and nothing here judges how good any of them is. Our own side is built to be checked the same way: published day rates, published engagement prices, an open-source engineering standard of 72 rules, and twelve engagements written up with the evidence basis stated on each. Then put the same questions to every firm on your list, ours included, and give most weight to whatever you could verify without the supplier's help.