Anthropic has raised more than seven billion dollars. It built Claude, one of the most capable AI systems on the planet. It has some of the best researchers in machine learning working in its offices.
It does not have a sales team.
Instead, Anthropic paid Accenture — through partnership economics — to train 30,000 consultants to deploy Claude into enterprises. Not to sell Claude. To deploy it. To take a general-purpose AI model and make it work inside specific businesses, with specific workflows, in specific regulatory environments.
Think about what that decision reveals. The company that built the technology concluded that the technology alone is not enough. That the gap between what AI can do and what a business needs it to do is too large for the model to cross on its own. That gap is not a sales problem. It is a knowledge problem.
The deployment problem nobody talks about
AI models are general-purpose by design. Claude can draft a contract, summarise a filing, analyse a dataset, write a compliance report. It can do all of these things competently. What it cannot do is know which of those things matters most to a specific firm, in a specific context, on a specific Tuesday morning.
A 200-person insurance broker has different priorities to a 30-person accountancy practice. A firm under FCA supervision has different requirements to one regulated by the SRA. The model does not know these differences. It cannot. They are not in the training data.
This is the deployment problem. The AI works. It just does not land. The distance between a working demo and a production system that changes how a firm operates is almost entirely made of domain expertise — the knowledge of how things actually work inside that specific business, in that specific sector, under those specific obligations.
Why Accenture, not a salesforce
Anthropic could have hired thousands of salespeople. The unit economics would have been straightforward: pay people to convince enterprises to buy API access. Standard SaaS playbook.
They did not do that. Because they understood something that most AI companies have not yet grasped: the sale is not the hard part. The application is.
Accenture brings the client relationships. The regulatory understanding. The industry-specific knowledge. The implementation experience. Accenture's 30,000 consultants are not evangelising AI to sceptical executives. They are configuring it for specific use cases, inside specific firms, against specific business problems.
Anthropic cannot do that. It does not have 30,000 people who understand insurance brokerage, or wealth management, or legal practice management, or pharmaceutical compliance. No AI company does. The domain expertise sits with the people who do the work — or with the consultants who advise them.
The pattern is not new
This is not an AI invention. It is a platform pattern as old as enterprise software.
Microsoft did not deploy Office 365 into enterprises alone. It built a partner ecosystem of hundreds of thousands of consultants who handled implementation, training, and ongoing management. Salesforce did the same. SAP built an entire industry of systems integrators around its platform.
Every platform technology that successfully reached enterprise did it through domain experts, not direct sales. The technology company builds the capability. The domain expert applies it. The client gets something that works for their specific situation.
AI follows exactly the same pattern, for exactly the same reason: the technology is general, but the application is always specific. Someone has to bridge that gap. That someone is never the technology company.
What this means at SME scale
A 30-person law firm is not going to engage Accenture. Neither is a 15-person accountancy practice, or a boutique wealth manager, or an estate agency with four offices.
But these firms have exactly the same problem. AI works in general. It does not work for their specific practice, their specific clients, their specific way of doing things — not without someone who understands that context and configures it accordingly.
The domain expert who bridges the gap between general AI and specific application is the same role at every scale. At enterprise scale, it is Accenture with 30,000 consultants. At SME scale, it is an individual consultant — an accountant who understands practice management, a compliance specialist who knows MLR obligations, a legal operations expert who has spent years inside firms like the one they are advising.
The role is identical. The scale is different. The need is the same.
The expert is not optional
This is the part that matters most. The domain expert is not a nice-to-have distribution channel. Not a convenient way to reduce customer acquisition costs. Not an add-on.
The domain expert is a structural requirement of making AI work.
Without someone who understands the firm's work, AI draws on public knowledge and produces generic output. The same output your competitor gets. The same structure, the same language, the same conclusions. Because it is working from the same pool of explicit, universally available information.
With someone who understands the work — who can capture the tacit knowledge, the judgment calls, the client intelligence, the operational nuances that make a firm what it is — AI produces something no competitor can replicate. Not because the model is better. Because the knowledge feeding it is unique.
Anthropic understood this. That is why they paid for domain expertise instead of hiring salespeople. The technology without the expertise is just a demo. The expertise without the technology is just consulting. Together, they are something neither can be alone.
The same logic, applied
itisyour.ai recognised the same structural reality. The platform captures a firm's knowledge — its constitution, its domain expertise, its operational intelligence — and makes it available to AI. But the platform alone does not land in firms any more than Claude alone lands in enterprises.
That is why itisyour.ai pays consultants to build on the platform. Not charges them. Pays them. The consultant brings the domain expertise. The platform provides the infrastructure. The firm gets AI governance built by someone who understands their work.
It is the Anthropic model at SME scale. Same logic. Same structure. Same recognition that the technology needs the human knowledge to be worth anything at all.