Book Your Call
One call. Clear direction. Faster execution.
How a 400-person company went from AI curiosity to a company-wide rollout in 90 days.
Leadership knew AI mattered but had no policy, no plan and staff quietly using personal accounts. Skillfi designed the whole setup, from policy and workspace architecture to seat allocation and usage limits, then trained every department. It's how our Claude workshops work: real work, real workflows, lasting habits.
Everyone was using AI. Nobody was managing it.
The leadership team knew AI would change how the company worked, but there was no policy, no budget owner and no plan. Meanwhile, a staff survey showed 41% of people were already using AI through personal accounts, sometimes with client data.
A first quote to give everyone the top-tier licence would have cost far more than the board would approve. Nobody knew which teams needed heavy usage and which needed light access, whether AI should be run centrally by IT or left to each department, or how to stop the early users from hitting usage limits by midweek.
Our people were already using AI. The only question was whether we'd manage it or keep pretending it wasn't happening.
COO, client
Design the whole setup first, then train every team on it.
We started with leadership, not licences. Policy, architecture and budget were agreed in an executive workshop, and only then rolled out department by department. Data and security rules were reviewed with our fractional CISO.
Central standards, department ownership
A hub-and-spoke model: a small central AI council owns policy, security and the platform, while champions in each department own use cases and training.
One workspace, built for every team
Single sign-on, admin controls and data rules, with shared Projects and prompt libraries for each department and company-wide ones for policies and brand.
The right seat for the right work
Seat tiers allocated by department based on real usage, plus training on how to work within usage limits, so heavy users stopped running out and light users stopped costing too much.
90 days, step by step
From scattered personal accounts to a governed, company-wide rollout. Here's how the engagement unfolded.
- Week 1–211 departments
AI readiness review
Staff survey, current AI use, data risks and use-case ideas collected from every department.
- Week 3Policy signed off
Executive workshop
AI policy, safe-use rules, central vs department ownership, budget and the 90-day roadmap agreed by leadership.
- Week 4–5400 seats right-sized
Architecture & seats
Workspace, sign-on, admin controls and department Projects set up, with seat tiers allocated by real need.
- Week 622 champions
Champions trained
Two champions per department trained to lead use cases, answer questions and coach their teams.
- Week 7–1122 use cases live
Department workshops
Hands-on workshops for every team on their own work, with two use cases per department taken live.
- Day 9074% weekly use
Board review
Adoption, usage, cost and results presented to the board, with seat allocation tuned and the next roadmap set.
Central rules, team ownership
The four layers that keep AI safe, affordable and actually used across 400 people.
Clear rules for everyone
- Safe-use & data rules
- Approved tools only
- Central AI council
One workspace, set up right
- Single sign-on & admin controls
- Department Projects
- Shared prompt libraries
Spend where it pays
- Seat tiers by department
- Usage limit best practice
- Quarterly right-sizing
Champions in every department
- Use cases owned locally
- Peer coaching
- Wins shared company-wide
Ninety days after kickoff
Staff using approved AI every week
And personal accounts with company data are now close to zero.
In 90 days we went from a risk we were ignoring to a capability we're proud of. Every team has AI that fits its work, and we're paying for what we actually use.
COO, clientIs AI happening in your company without a plan?
We'll run the executive Claude workshop, design your AI setup from policy to seats and train every team on their real work. Need security and data rules in place first? Bring in a fractional CISO, and a fractional CTO for the wider tech roadmap.
- An AI policy & 90-day roadmap
- Workspace architecture & governance
- Seat allocation & usage optimisation
- Workshops & champions in every team
30 minutes. No obligation.
Related Case Studies