Programme · AI Capability Foundation

AI transformation programme: from pilots to an AI operating model you run

Stop running AI pilots. Start running an AI company.

AI Capability Foundation brings together what companies usually buy separately: an interim Head of AI who owns the outcome, a roadmap tied to your P&L, the first use cases in production, training for leaders and staff, and governance built into delivery. One accountable owner, 6 to 12 months, and at the end your own successor runs it without us.

Duration 6-12 monthsOwner one, namedPrice fixed per phaseEnds with your own successor

Book a free 30-minute call See the four phases

Who this is for

Companies from 50 employees, from owner-managed firms to international groups, with an AI ambition and no internal operating model to sustain it.
CEOs, CTOs and CDOs asked by the board for AI results and by regulators for AI control (EU AI Act Article 4, GDPR).
Organisations where the next permanent AI hire is still months away.
Companies with several pilots, a half-finished strategy and no single person who owns the outcome.

1,000+employees enabled

110+AI Champions developed

12markets

The enablement model in this programme was built and run by our founding team in a senior role at a European electronics group. Read the case study

What you get

Four pieces, one owner

Each piece also exists as a service of its own. In the programme they are planned, staffed and steered together, so nothing waits for anything else.

Interim AI leadership and roadmap

We take the interim Head of AI seat: we run the weekly cadence, chair the steering committee, keep the roadmap honest, and hire and onboard your permanent successor. The roadmap ties two or three use cases to real P&L outcomes and sequences them against data readiness, capacity and regulatory risk.

A prioritised 12-month roadmap with a business case per initiative
A decision log of what is explicitly out of scope
Steering committee and weekly cadence from week one
Recruiting and onboarding of your internal successor

Service: Fractional AI Leadership →

First use cases in production

Two use cases go live in the pilot wave, built by our engineers in your environment, with monitoring and an owner on your side. Everything else in the programme is built around running systems, not slides.

Two use cases live within about five months
In your cloud or on your premises; your data stays yours
Monitoring, retraining and handover documentation
A second wave planned on what the first one taught

Service: Production AI →

Enablement at three levels

Leaders, staff and functions learn at the same time, because training at one level alone does not stick.

Leaders: a course on portfolio, risk and trade-offs, not a technology course
All staff: AI literacy pathways, documented as evidence for Article 4 of the EU AI Act
Functions: role-based tracks, hackathons and an AI Champions network
Materials your team keeps and runs after we leave

Service: AI Enablement & EU AI Act Readiness →

Governance built into delivery

Legal, data protection, the works council and risk sit in the room when data and model decisions are made, not only at the final review. That is faster, and the result holds up.

AI inventory and EU AI Act risk classification
Review gates in your delivery process, with named approvers
The works council involved from the start (section 87 BetrVG)
Explainability and logging designed in before go-live

Guide: the EU AI Act for CTOs →

Why the four pieces belong together

Bought on its own	What you get	What is missing
Strategy	A roadmap	Someone to deliver it, and a way to ship through governance
Interim leadership	An experienced person in the seat	An agreed roadmap and a trained team to lead
Enablement	Trained people	Live use cases to apply it to, and clearance to ship
Governance	Policies	Priorities to govern; without them the business works around the rules

If you only need one of these, book that service. If you need all four to work, book the programme.

Four checkpoints before anything goes live

AI pays off when two things are true at once: people are able to use it, and decisions about it are made well. The programme builds both. Every output that reaches a customer, an employee or a regulator passes four checkpoints, built into delivery from day one.

Context

What does the model not know that a person would weigh?

Challenge

When does a result get a second look instead of a rubber stamp?

Choice

Which trade-offs are acceptable, and who signs off on each one?

Accountability

Who owns the final decision? A named person, not a committee.

Path to production

Four phases, 6 to 12 months

Weeks 1-6

Foundation

Assessment of skills, decisions and governance
12-month roadmap with business cases
Governance gaps mapped against the EU AI Act
Interim leader in place, steering committee running

Months 2-5

Pilot wave

First two use cases in production, with the four checkpoints
First AI Champions cohort trained
Review gates live in the delivery process
First hackathon with working prototypes

Months 5-9

Scale

Second wave of use cases
Leaders programme delivered
Role-based tracks across business units
Compliance checks built into tooling, not a ticket queue

Months 9-12

Handover

Internal successor confirmed
Successor runs the cadence for 90 days, with us in the background
We step out; a quarterly check-in if you want one

Three ways to start

6 weeks fixed price

Diagnostic and roadmap

Assessment, governance gap analysis against the EU AI Act, a 12-month roadmap and a proposal for the interim lead. The smallest way in.

6-9 months fixed price

Capability build

The programme through the pilot wave: first use cases live, first leaders and Champions cohorts trained, governance gates working.

9-12 months fixed price

Full foundation

End to end, through handover. Your successor runs the operating model for 90 days before we step out.

Fixed price per phase, agreed before it starts. Capacity-limited to two active clients at a time.

Solution, programme or service?

Solution

When a well-known problem needs a proven answer

A packaged approach with a fixed scope
Demo, pilot, then a monthly run
For example ProDataAI Lens video analytics

Explore →

Programme

When you need the AI operating model built

Leadership, use cases, enablement and governance together
One accountable owner for 6-12 months
Ends when your own successor runs it

Service

When you have an owner and need one specific skill

One gap filled: leadership, enablement or a build
Scoped together with you
From a few weeks to several months

Explore →

Questions about the programme

What is an AI operating model?

The way your company decides on, builds, runs and controls AI: who owns AI initiatives, how use cases are chosen and approved, who is trained for what, and how governance works in daily delivery. AI Capability Foundation builds it with you and hands it over to your own team.

How long does an AI transformation programme take?

Six to twelve months. The first six weeks produce the assessment and the roadmap, the first use cases go live in months two to five, and the handover follows in months nine to twelve, once your internal successor has run the cadence for 90 days.

Do we need a Head of AI before we start?

No. We fill the interim Head of AI role from day one and help you hire and onboard the permanent person. The programme ends when that person runs the operating model without us.

Does the programme cover AI literacy under Article 4 of the EU AI Act?

Yes. Article 4 has applied since February 2025 and requires companies that use AI to ensure their staff have sufficient AI literacy. The programme delivers role-based training for leaders, staff and functions and documents it as evidence. It is not legal advice: your legal team keeps the final say.

How is this different from an AI strategy project?

A strategy project ends with a roadmap. This programme ends with use cases in production, trained people, governance that works in daily delivery, and your own person in charge. The roadmap is week six, not the finish line.

Is the works council involved?

Yes, from the start. Where AI systems could be used to monitor employees, section 87 BetrVG gives the works council a say. We involve it in the governance design and in use-case decisions, which avoids delays at go-live.

What does it cost?

Each phase has a fixed price, agreed before it starts, based on the size of your company and the scope you choose. You can start with the six-week diagnostic and decide on the rest afterwards.

Ready for a 30-minute conversation?

Tell us what your AI capability looks like today, who owns it and where the gaps hurt. We will tell you honestly whether the programme fits, or which single service is the better start.

Book a free 30-minute call

or email ai@prodata.ai