Stage 3 · Run

Managed AI Service & Hosting

Once something is built, someone has to keep it working. We'll do it for you, or show your team how.

Once something is built, someone has to keep it working. The AI models it uses get retired and replaced, connections break when a supplier changes its software, and your own process changes over time. Keeping on top of that is steady, ongoing work, and most teams don't have spare time for it. We'll do it for you, or show your team how.

Where It Runs

Your decision

OptionHow it worksSuits
Your environmentWe build into your own Azure, AWS or Microsoft 365 tenancy. You host it. We provide support and iteration.Businesses with an established platform or IT team
Our managed environmentHosted in a UK region, isolated per client. We run, monitor, patch and maintain it. Nothing for your team to set up or support.Businesses without spare technical capacity
HybridYour data stays in your environment and the software that co-ordinates the work runs in ours. We agree which part sits where, piece by piece.Businesses with specific data residency or control requirements
What's Included

The managed service

Platform & security

  • Hosting in a UK region, kept separate from other clients' systems
  • Sign-in, access control, and passwords and keys stored securely
  • Backup, retention and disaster recovery
  • Defined availability target and service levels

Operations

  • Monitoring and alerts on whether it's running, how much it's doing and how often it fails
  • Patching and platform upgrades
  • Incident response within agreed service levels
  • Cost monitoring, using cheaper AI models wherever they do the job well enough

Keeping pace with the models

  • Moving to new AI models when providers release them or retire old ones
  • Every change tested against a set of your own examples before it goes live, so any difference in results is caught first
  • Tuning as your process, templates or rules change

Assurance

  • A log of what the system did, and records you can show your accountant, auditor or insurer
  • Monthly usage, cost and exception reporting
  • A quarterly review against your original ranked list
  • Included development hours every month for adjustments and tuning
Compliance & Data Protection

Your data and controls stay protected

  • UK data residency, with data flows documented and mapped
  • Data processing agreement with named sub-processors, and how data would leave the UK if it ever needs to
  • Help with your data protection impact assessment (DPIA) and records of processing (ROPA)
  • Retention, deletion and audit trail design agreed with you in advance
  • Documentation prepared for you, so you aren't left assembling it when a customer, insurer or auditor asks
Exit Terms, Published Up Front

How leaving works

  • A clearly stated notice period
  • Transition support over 30, 60 or 90 days, priced at the outset
  • All the code, settings and set-up scripts handed over in full
  • A full export of your data in documented, standard formats
  • Practical help setting it up in your own Microsoft, Azure or AWS account

We want you to stay because the service is worth paying for, not because leaving is difficult.

Pricing

Priced individually based on the systems in production, your chosen hosting option, and the service levels you need. Model and infrastructure costs are always passed through at cost, itemised monthly, with no hidden margin on usage.

Related Services

You might also need

Let's talk about Managed AI Service & Hosting

Start with a free 30-minute scoping call. We'll tell you whether a managed service is the right first step.

Book a free call