# AI Transformation

- Link: https://xanadu.co/services/ai-transformation/
- Published: 2026-07-16T19:27:00+04:00

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.

[Take the free assessment](https://xanadu.co/ai-readiness/)

[Book a scoping call](https://calendar.superhuman.com/book/11VUS7a5FE0mTLn5Oa/sMOZL)

![An organisation-wide AI opportunity heatmap ranking functions by value and readiness](/
wp-content/uploads/2026/07/xa-heatmap-ai-hero-v3.png)

The journey

## One journey, five steps.

Sold one at a time. Each step ends with the reason for the next.

 0. [Step 1AssessFree, 3 minutes. Know where you stand.](https://xanadu.co/services/ai-transformation/#assess)
 1. [Step 2Blueprint3 to 4 weeks, fixed fee. The plan, the licence case, and a date.](https://xanadu.co/services/ai-transformation/#blueprint)
 2. Scoped off the Blueprint roadmap
 3. [Step 3Enable4 to 6 weeks. The go-live: deployed, launched, trained, adopted.](https://xanadu.co/services/ai-transformation/#enable)
 4. [Step 4Govern3 weeks. Risk and spend under control as usage scales.](https://xanadu.co/services/ai-transformation/#govern)
 5. [Step 5Build8 to 12 weeks per function. Agents in production, off your roadmap.](https://xanadu.co/services/ai-transformation/#build)

[Executive AI Coaching](https://xanadu.co/services/ai-transformation/#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.**

[Take the assessment](https://xanadu.co/ai-readiness/)

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

Duration3 to 4 weeks.

CommercialsFixed fee, agreed before we start.

You bringAn executive sponsor and access to the people who do the work.

Not a strategy deck. A plan with a date on it.

![A Claude licence plan and total cost of ownership model](/wp-content/uploads/2026/
07/xa-blueprint-tco.png)

[Book a scoping call](https://calendar.superhuman.com/book/11VUS7a5FE0mTLn5Oa/sMOZL)

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

Duration4 to 6 weeks to launch, plus the 30-day readout.

CommercialsFixed fee.

You bringYour 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.

![An adoption curve from a 30-day post-launch readout tracked against target](/wp-
content/uploads/2026/07/xa-enable-adoption-v2.png)

[Book a scoping call](https://calendar.superhuman.com/book/11VUS7a5FE0mTLn5Oa/sMOZL)

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

Duration3 weeks.

CommercialsFixed fee.

You bringAn 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.

![A shadow AI usage baseline showing unsanctioned AI tools in use across an organisation](/
wp-content/uploads/2026/07/xa-govern-shadow-ai-v3.png)

[Book a scoping call](https://calendar.superhuman.com/book/11VUS7a5FE0mTLn5Oa/sMOZL)

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.

![A reference architecture for a Claude agent deployed into a customer environment](/
wp-content/uploads/2026/07/xa-build-architecture.png)

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

Duration8 to 12 weeks per function.

CommercialsFixed fee per deployment.

DeliveryWith certified engineering specialists matched to your technology stack.

You bringSystem access and a named technical contact.

One function automated in production, and a proven pattern that makes the next one
faster.

[Book a scoping call](https://calendar.superhuman.com/book/11VUS7a5FE0mTLn5Oa/sMOZL)

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

Format5 sessions, 1:1, delivered personally by our founder.

DurationApproximately 10 weeks.

AvailabilityDeliberately limited.

![Daragh McGrath, Founder and CEO of Xanadu](/wp-content/uploads/2025/10/DM-Profile-
Business.webp)

[Book a scoping call](https://calendar.superhuman.com/book/11VUS7a5FE0mTLn5Oa/sMOZL)

## How we work with Anthropic

Xanadu is a Certified Services Partner, Claude Partner Network.

Canva handles creation and brand control. Claude handles the reasoning and automation
around it. We connect the two, then govern how they run across your organisation.

The operating model we use is published in full as [The AI Content Supply Chain](https://xanadu.co/ai-content-supply-chain/).

![Certified Services Partner, Claude Partner Network](/wp-content/uploads/2026/10/
claude-partner-network-certified-badge.svg)

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.

[Explore Canva →](https://xanadu.co/services/canva/)
