A base fee, plus what
the agents actually do

The base fee

Deploying an agent is not the same as buying a licence and hoping. The base fee covers everything that makes the agent succeed in your environment: onboarding, integration into your systems of record, configuration of the trust layer, monitoring, maintenance and ongoing support. When something in your source data changes, we handle it. When an agent needs retuning against new cases, that is included.

This is deliberate. Between 70 and 80 per cent of enterprise AI projects never reach production, and almost none of them fail for lack of a model. They fail because nobody owned the integration, the tuning and the day-two operation. The base fee is how we take that risk off your side of the table.

Per-task pricing

On top of that, agents are priced by the work they do rather than by seats provisioned. A month with more campaigns, more enquiries or more market analyses costs more; a quiet month costs less. Cost tracks the value being produced, which is what makes the business case legible to a CFO.

Why not per seat

An agent does not occupy a seat. Pricing it as though it did would mean charging you the same whether it ran once or ten thousand times, and that is not how a digital employee should work.

Getting a number

The specific base fee and per-task rate depend on which agents you run, which systems they touch and what volume you expect. We work these out during scoping and give you the full cost model before anything is built.