We run your
operations.

It is an arrangement in which a provider runs a business function for you and uses AI agents to do part of the work, with people handling exceptions, judgment and accountability. You buy the operating result, measured through service levels, rather than software licences or staff hours, and the provider remains responsible for improving the service over time.

Decisions that define your risk appetite or need your own accountability, such as credit policy, clinical judgment or final complaint outcomes, plus a small retained team: a service owner, process experts and control owners. Preparation, routine processing and evidence gathering around those decisions are usually the work that moves to the service.

Yes. Outsourcing moves the work, not the accountability, and in regulated sectors the law says so explicitly. That is why the contract should give you audit access, change approval for anything that alters outcomes, incident notification and an exit plan, and why your governance forums should review quality samples rather than only speed.

Yes, and it is often the better route. A single, well-measured process such as invoice processing or customer support triage gives both sides a baseline, a working governance rhythm and evidence about automation coverage. Further processes can then be added through the same six phases with far less uncertainty.

It depends on the design and the jurisdiction. Some people move to the provider, sometimes under staff-transfer law such as TUPE in the UK; some form the retained team that oversees the service; some move to other roles. Consultation should be honest about how automation is expected to change the work over time.

On outcomes rather than per full-time-equivalent person, as our comparison pages state. The exact structure, such as a base fee, unit pricing or a gainshare, is agreed per engagement once discovery has produced a baseline. We do not publish price lists, because the starting point of each process differs too much.