AI tools and platform rollout
The fastest value in most businesses is a governed rollout of tools people already use
In every discovery programme we run, the loudest frustration is the same: AI tools are blocked, so staff use personal accounts. We replace the block with sanctioned, governed access to Claude, Microsoft Copilot or ChatGPT Enterprise, selected on your evidence, and we make it stick.
We select between Microsoft, Anthropic and OpenAI on your evidence
The three enterprise assistants differ in where they live (inside Microsoft 365, in a standalone workspace, in your own applications), in their data terms, in how they handle documents and code, and in cost per seat. The right answer depends on what your people actually do all day and what systems hold your knowledge. We run a short, evidence-based selection: a sample of real tasks from your functions, tried on each, with the licence terms laid side by side.
We are platform-agnostic: we recommend Claude, Copilot or ChatGPT Enterprise on what your tasks and your data terms show, we take no margin on any licence, and we will say if the answer is two of them rather than one.
Governed, phased, measured
Licence and data terms
Enterprise tiers with contractual no-training commitments, data residency confirmed, and the terms written into your acceptable-use policy so IT governance can sign off.
Data classification and DLP
Clear rules on what may and may not be shared, backed by data-loss- prevention controls, so sensitive material is protected whatever the tool.
Role-based, phased access
A sanctioned-tool list, access opened by role and use case rather than all at once, an approval route for new tools, and a pilot group that becomes the champions.
Custom GPTs, Claude skills and projects
Configured assistants for the recurring work in each function: proposal drafting, meeting summaries, code review, customer replies, with your templates and tone built in.
Developer tooling
Claude Code, GitHub Copilot and Codex for engineering teams: setup, coding standards, review gates, token budgets and security, from a team that ships with these tools daily.
Adoption measurement
Usage by role and function, time saved on the sampled tasks, and the feedback loop into training, so the rollout proves itself within a quarter.
Weeks, not quarters, to a governed pilot
Selection takes two weeks, rollout design two to three, and pilot groups are usually live within six weeks of the start. Licence costs are paid to the vendor directly; we take no margin on any of them.
About tool rollouts
Claude, Copilot or ChatGPT?
It depends on where your work lives. Copilot wins when the work is in Microsoft 365 and Teams; Claude when the work is documents, analysis, code or building agents and the quality of reasoning matters most; ChatGPT Enterprise when breadth of integrations and consumer familiarity count. Many businesses run two. We test on your tasks rather than argue in the abstract.
Is it safe to let staff use these tools?
With enterprise licences, data classification and a usage policy, yes, and it is safer than the alternative. A blanket block does not stop AI use; it pushes it into personal accounts the business cannot see or control.
Do we need training?
Yes, and the EU AI Act has required AI literacy since February 2025. Our training and enablement service runs alongside every rollout.
How long does a rollout take?
Selection two weeks, design two to three weeks, then a phased rollout over one to two quarters depending on size. Pilot groups are usually live within six weeks of the start.
Replace the block with a plan
Tell us which tools are in use today, officially and unofficially, and we will come back with a selection and rollout approach.