Most businesses do not have an AI problem. They have a workflow problem.
We have run managed IT and security for creative and growing businesses since 2003. We find the workflows where AI earns its place, then build and run them safely.
Discovery first, pilot before scale
It starts with a free conversation. If it is worth going further, four stages follow, each one of our four services. The order shows what you get at each point.
AI Discovery Call
A 30 to 45 minute conversation with your decision makers to understand the business, its functions and its systems, and to work out whether an Assessment is worth doing. It is free, there is no written deliverable, and nobody is asked to commit to anything.
What we cover: the business functions and the systems in scope, the number of sites and users, and the stakeholders who should be in the room.
Stage 1: AI Readiness Assessment
We map the workflows, the systems and the data before scope hardens. You get ranked opportunities with baseline numbers, a policy and governance review, and a costed roadmap.
Stage 2: Secure AI Deployment
The chosen AI tools rolled out with identity, DLP and browser controls in place first. You get a stack your staff can use without putting client work into consumer AI.
Stage 3: AI Workflow Build
One workflow, built and run in parallel with the existing process. You get a working pilot measured against agreed criteria, and a go/no-go decision point before any scale-up spend. If the numbers are not there, we stop, and you have lost one pilot rather than a programme.
Stage 4: Managed AI
Production rollout with runbooks, monitoring and human gates. You get value reporting at 30, 60 and 90 days, then a quarterly AI review.
The pilot gate is a real gate. If the pilot does not prove out at Stage 3, you can stop there, before any scale-up spend.
From assessment to managed AI
Start with the Assessment. Everything after it is scoped and fixed-priced from what it finds.
AI Readiness Assessment
Fixed price, agreed before work starts.
Workflow discovery and audit, ranked automation opportunities with baseline numbers, shadow AI audit, policy and contract-clause review, staff enablement session, costed roadmap. Typically two to three weeks, and around half a day of your team's time. 25% of the fee is credited against a Secure AI Deployment project won within 90 days.
Secure AI Deployment
Scoped and fixed-priced from the Assessment.
Copilot or Claude rolled out with SSO, DLP and browser controls so staff cannot paste client work into consumer AI, plus identity hardening.
AI Workflow Build
Scoped and fixed-priced from the Assessment.
Custom automation and agent builds. Pilot first: one workflow, run in parallel, prove it, expand.
Managed AI
Priced on what is under management.
Licence management, usage reporting, policy refresh, quarterly AI review, and keeping every deployed workflow doing what we agreed. Covers AI systems built by someone else, adopted after an onboarding audit.
How pricing works
- Every project is quoted as a fixed price, agreed before work starts.
- Payments are staged against objective milestones.
- Expected monthly running costs are estimated in the proposal.
- No hidden costs and no open-ended billing.
What the business gets back
Capacity without headcount
The repetitive work moves to a system that does it the same way every time, and your team keeps the judgment.
Hours returned, and measured
We capture the baseline before we build, then report the hours recovered at 30, 60 and 90 days, so the gain is measured rather than asserted.
Errors removed from routine work
Missed chases, re-keyed data and inconsistent handling cost real money. Automated steps remove the variation and a person still approves the exceptions.
Institutional knowledge made usable
Contracts, proposals, project history and process documents become answerable with sources, inside your own environment, so what one person knows becomes something the business can query.
The value case comes before the invoice
Every proposal sets out what the problem costs you per year, in your own numbers, against our fixed fee. If those two figures do not justify each other, we will say so.
Change control you already recognise
Every AI change is reviewed, documented and reversible. It is the same discipline we apply to the rest of your systems, and it is why AI work does not become the thing nobody can unpick.
You own your deployment: your data, your configuration, your instance, your keys. You keep all of it, whatever happens between us.
Where AI and automation pay off
An outline of the areas where AI earns its place. A map of where to look, not a menu to order from.
Finance
- Credit control: invoices chased to paid or escalated, politely and consistently, with a human approving the exceptions.
- Invoice and expense processing: supplier invoices and receipts extracted, coded and routed for approval automatically.
- Month-end reporting: the management pack drafted from your accounting data, ready for review.
Sales and marketing
- Lead qualification and outreach: companies discovered, scored against your ideal profile and fed into your CRM. This is how our own new-business engine runs today.
- Proposal and tender drafting: first drafts built from your past wins, rates and scope notes.
- Content production: case studies and posts drafted from real project work, in your brand voice.
- Meeting notes to actions: transcripts turned into summaries, actions and CRM updates.
Client service and delivery
- Client onboarding: accounts, folders, checklists and welcome communications created from a single trigger.
- Inbox and ticket triage: incoming email classified, routed and drafted for human review.
- Client support assistant: client questions answered from your own documents, with a human one step away.
Knowledge and assets
- Contract review: contracts and project documents made answerable, with sources, inside your own environment.
- Company knowledge assistant: your handbook, processes and guidelines, queryable by your team.
- Asset tagging and search: media libraries automatically tagged so your archive becomes findable.
People and operations
- Staff onboarding and offboarding: accounts, licences and devices set up on day one and shut down on departure.
- CV and applicant screening: shortlists against your stated criteria, with people making every decision.
Why your IT partner should run your AI
We already hold the permissions model
AI tools inherit whatever access the underlying systems grant, so someone has to check who can see what before anything is switched on. We already manage the identity and file permissions, so that check is work we can actually do.
The governance layer bolts onto what we already run
DLP, browser controls, audit logs and monitoring sit on the identity, device and security stack we manage for you already, rather than being bolted on by a third party who cannot see it.
We work with the tools you already own
Most of the value is in software you are already paying for. We would rather switch on and govern what you have than sell you another platform.
We run this on ourselves first
- Our own new-business engine runs on an AI agent we built: company discovery, scoring against an ideal profile, and outreach into our CRM.
- We pilot automations on our own service desk before we offer them to clients.
- We apply the same change control to our own AI work that we apply to client systems.
Questions we get from cold enquiries
Book a free AI Discovery Call
A 30 to 45 minute conversation with your decision makers to understand the business and work out whether an Assessment is worth doing. Free, no written deliverable, no commitment.