The reasoning layer
AI that sticks.
Most organisations have experimented with AI. Very few have made it part of how they operate. The difference is not the model. It is the work between the model and your business, and that work is what we do.
We take organisations from AI experimentation to Claude running in production: deployed, adopted, governed, and automated into the workflows that run the business.
The journey
One journey, five steps.
Sold one at a time. Each step ends with the reason for the next.
- Step 1Assess
- Step 2Blueprint
- Scoped off the Blueprint roadmap
- Step 3Enable
- Step 4Govern
- Step 5Build
Executive AI Coaching runs alongside any step, for the leader driving it.
Every engagement is fixed fee, fixed scope, and stands alone.
Senior operators, small teams, no bench.
Step 1 · Free
AI Readiness Assessment
Free, 3 minutes. Know where you stand.
Step 2 · Paid
The AI Blueprint
3 to 4 weeks, fixed fee. A plan with a date on it.
You have decided AI matters. The question now is where it pays, what it costs, and how to start without betting the business on a guess. The AI Blueprint answers all three in under a month.
We scan the whole organisation for where AI creates value, then go deep on the function where it pays fastest: how your people make decisions, create documents and content, find what they need to know, analyse information, and run the repeatable processes that eat their week.
What you walk away with
- The opportunity heatmap. Organisation-wide and prioritised, so one Blueprint shows the whole journey.
- A first-function deep-dive. Every valued task in your first function mapped to a specific Claude recommendation.
- The blockers, named honestly. Each with a remediation plan.
- A right-sized licence plan and TCO model. Who needs which licence, expected consumption, and what it costs at your scale — no overbuying, no surprise bills.
- Adoption targets. Set before anything is deployed.
- A fundable business case. One your CFO can underwrite.
- A go-live date and a named project team. The two things most AI strategies never produce, with every role identified.
Why one function deep, not twenty shallow
Depth is what makes a plan executable. The heatmap shows where the rest of the value sits, and the method repeats across functions, faster each time, without repeating the price of the first.
The detail
- Duration
- 3 to 4 weeks.
- Commercials
- Fixed fee, agreed before we start.
- You bring
- An executive sponsor and access to the people who do the work.
Not a strategy deck. A plan with a date on it.
Step 3 · Launch
Enable
4 to 6 weeks. The go-live.
Blueprint is the strategy. Enable is the launch. Most AI deployments fail quietly: licences get bought, a pilot impresses, usage fades within a quarter. Nobody announces the failure. It just stops coming up. A proper launch is how you avoid being one of them.
Enable is the engagement that hits the date your Blueprint set: the technical deployment, the project, the change campaign and the training that turn a plan into a launch, and a launch into daily use.
What you walk away with
- The technical deployment. Claude provisioned and configured to your licence plan: SSO and identity integration, user provisioning, permissions, and security and data controls set to your policy. Core connectors switched on for the tools where work already happens, so Claude starts useful, not empty.
- A managed go-live. Kickoff, named workstream owners, weekly cadence, risk and dependency tracking, and a launch readiness review before the date.
- A launch your people notice. Pre-launch comms, early access for leaders and champions, a go-live moment, and a post-launch drumbeat that keeps usage building. Plus an executive visibility playbook, because adoption follows what leaders are seen to do.
- Training that lands. Role-based sessions by audience: executives, champions and teams, built around their actual workflows. Prompt libraries from their real tasks, with recorded assets and guides that outlive the sessions. Named champions in every function.
- Adoption, proven. Tracked against the targets set in your Blueprint, with a 30-day readout and a handover your team can run without us.
The detail
- Duration
- 4 to 6 weeks to launch, plus the 30-day readout.
- Commercials
- Fixed fee.
- You bring
- Your executive sponsor, your named champions, and IT access for identity and configuration.
When usage is real, two questions arrive on their own: is this under control, and what does it cost at scale. We answer both.
Step 4 · Control
Govern
3 weeks. Risk and spend under control.
Your people are already using AI. The only question is whether you can see it. In almost every organisation we assess, AI is in daily use through consumer accounts, personal logins and tools nobody sanctioned, with company data moving through all of it. And alongside the risk sits the other fear that stalls every AI programme: the cost blowout nobody saw coming.
Govern brings all of it, risk, spend and shadow usage, inside a framework your board can sign.
What you walk away with
- A usage baseline. Every AI tool in use across the organisation, sanctioned or not.
- A data exposure inventory. What is going where.
- AI spend, visible. What AI actually costs you today, across every tool, seen in one place.
- Cost guardrails. Budget controls, so consumption scales with value and a blowout is designed out rather than discovered on an invoice.
- A governance workshop. With leadership, IT and legal.
- An AI usage policy. Drafted to your context, your sector and your jurisdiction.
- A data governance framework. Practical guardrails your teams can follow.
- A board-ready readout. With a remediation plan.
The detail
- Duration
- 3 weeks.
- Commercials
- Fixed fee.
- You bring
- An executive or CISO sponsor, and a willingness to hear an honest number.
Also available standalone. Many organisations start here, after a board question or a regulatory prompt.
A governed estate is a deployable estate.
Step 5 · Production
Build
8 to 12 weeks per function. Agents in production. Not a pilot.
The industry is full of impressive demos. Build is for the other thing: agents and workflows doing real work, in production, integrated into the systems your business already runs on.
Your heatmap has more than one line on it. Build works through it, function by function.
What you walk away with
- A reference architecture. Built for your environment.
- Systems integration. Into the systems where your data already lives.
- An agent in production. Built, tested and deployed with real users.
- A runbook and full handover. Plus 30 days of hypercare.
It is yours: your platform, your data, your outcomes.
The detail
- Duration
- 8 to 12 weeks per function.
- Commercials
- Fixed fee per deployment.
- Delivery
- With certified engineering specialists matched to your technology stack.
- You bring
- System access and a named technical contact.
One function automated in production, and a proven pattern that makes the next one faster.
Alongside any step
Executive AI Coaching
For the leader the board is asking.
The board is asking about AI, and the honest answer in most executive teams is that nobody at the table uses it seriously. This programme fixes that for one person: you.
Five private sessions across ten weeks. We move you onto Claude personally, build your own agents around how you actually work, and embed AI into your daily operating rhythm. By the end you are not sponsoring the AI agenda. You are running it.
The detail
- Format
- 5 sessions, 1:1, delivered personally by our founder.
- Duration
- Approximately 10 weeks.
- Availability
- Deliberately limited.
The creative layer
The other half
Claude is how your organisation reasons. Canva is how it creates. Together they are the chain from prompt to presentation.