AI adoption & implementation

From individual experiments to a team that builds for itself.

You have the licences. Some of the team are experimenting enthusiastically; others do not know where to start. I help turn this into one shared approach: from discovery to working agents and workflows, with people who can take them further themselves.

An agent is an AI assistant that carries out a task in steps.

Thierry smiling during an AI presentation

A selection of organisations I have worked with

Through my own practice and previous roles.

  • Philips
  • Pathé
  • Woonstad Rotterdam
  • Exact
  • Basic-Fit
  • Yuki
  • Rabobank
  • de Jong Intra Vakanties
View all 21 references
  • Rabobank
  • Philips
  • PostNL
  • Yacht
  • DELA
  • ARAG
  • Pathé
  • ABN AMRO
  • a.s.r.
  • de Jong Intra Vakanties
  • Woonstad Rotterdam
  • IkBenFrits
  • Bright Pensioen
  • Exact
  • Twikey
  • Techleap
  • Yuki
  • Basic-Fit
  • ANWB
  • Financial Lease
  • RELX

Where things get stuck

  • Some people are hesitant; others are experimenting without coordination.

  • Pilots that never get beyond the demo and become part of everyday work.

  • Only two or three people know what is and is not allowed with data and AI.

Participants talking during a workshop with Thierry

People, processes and technology

With AI, 70 per cent of the work is about people and processes.

  • People

    The why and the question “what will be left of my profession?” determine whether people start building.

  • Processes

    Work that genuinely changes, with owners and ground rules.

  • Tech

    The simplest form that works, in the environment you already have. With tests defined in advance.

BCG uses 10-20-70: 10% algorithms, 20% technology and data, 70% people and processes. A practical rule of thumb, not a measured law.

Where adoption falters

Training addresses two of the five steps.

I use ADKAR: five steps a person goes through before a new way of working sticks. The first step that falters tells you where to intervene.

Five steps in AI adoption, from awareness to reinforcement.
 Awareness DesireThis is where training fitsKnowledgeThis is where training fitsAbility Reinforcement
What you hear“Another new tool. Why, exactly?”“What does this mean for my profession?”“I want to, but where do I start?”“It worked in the training. Not on Monday.”“After three weeks, everyone is doing things the old way again.”
What is neededMeaning: why this matters to our goalAutonomy and mastery: a say in how your work changesRelevance: knowledge applied to your own tasks, beyond a generic demoTime and colleagues: practising together on real workA system that rewards the new behaviour
What I doKick-off and team workshop on why we are doing thisInterviews and a first demo using your own tasksWorkshops using your own tasks and use casesTraining on real work, peer learning and sharing what goes wrong tooAdjust roles, routines and recognition together with managers
What you getCore narrative and baseline assessmentTeam-level diagnosisUse case backlogProven ways of workingReinforcement plan and follow-up assessment

Awareness

What you hear
“Another new tool. Why, exactly?”
What is needed
Meaning: why this matters to our goal
What I do
Kick-off and team workshop on why we are doing this
What you get
Core narrative and baseline assessment

Desire

What you hear
“What does this mean for my profession?”
What is needed
Autonomy and mastery: a say in how your work changes
What I do
Interviews and a first demo using your own tasks
What you get
Team-level diagnosis

Knowledge

This is where training fits

What you hear
“I want to, but where do I start?”
What is needed
Relevance: knowledge applied to your own tasks, beyond a generic demo
What I do
Workshops using your own tasks and use cases
What you get
Use case backlog

Ability

This is where training fits

What you hear
“It worked in the training. Not on Monday.”
What is needed
Time and colleagues: practising together on real work
What I do
Training on real work, peer learning and sharing what goes wrong too
What you get
Proven ways of working

Reinforcement

What you hear
“After three weeks, everyone is doing things the old way again.”
What is needed
A system that rewards the new behaviour
What I do
Adjust roles, routines and recognition together with managers
What you get
Reinforcement plan and follow-up assessment

Based on ADKAR (Prosci), supplemented with my own working model.

Sources

Model: ADKAR (Prosci). Change succeeds for each person through five successive steps. The first step that falters tells you where to intervene. Supplemented with Drive (Daniel Pink), Bas Kemme (IntotheNXT, 2026) and the 4D AI Fluency Framework by Rick Dakan and Joseph Feller (CC BY-NC-SA 4.0). The connection is my own working model.

4D dimension

Awareness
None: the “why” is a leadership narrative
Desire
Delegation
Knowledge
Description
Ability
Discernment
Reinforcement
Diligence

The statements are typical examples, not client quotes.

Book an adoption assessment

How I work

How we build AI adoption

  1. Discovery

    Interviews, workshops and a team profile. A catalogue of opportunities from across the organisation, beyond a single enthusiastic user.

  2. Ground rules

    What must AI not do? Who decides? Where does the data stay? Your IT lead's boundaries are the starting point.

  3. Build

    Assistants, agents and workflows. The simplest form first, with tests defined up front and short feedback cycles.

  4. Train

    Guides and support. Team members who start building their own agents.

I work with the software you already use. Want an AI environment set up with your organisation’s knowledge and clear ground rules? Tisser offers that. I co-founded the company and now work with it as an implementation and adoption partner.

  • MS Copilot (365, Cowork and Agent Studio)
  • ChatGPT / Codex
  • Claude (Code)
  • Tisser.ai

Related work

  1. AI adoption · Investment firm

    Getting the whole team working with AI (NL)

    Some were hesitant, others were running isolated AI experiments. I helped the team develop a shared approach to Microsoft Copilot and AI, with ground rules and initial applications. Three agents are now running, and colleagues are building new ones themselves.

    Thierry presenting AI to a group around a table
  2. Through an agency · anonymous

    Mapping AI opportunities for a family business (NL)

    Eight interviews, 31 use cases with an owner and a route, a workshop and a keynote.

    Thierry presenting an AI programme to participants

AI workshop with Thierry

“Thierry made AI tangible for our work as an agency. With humour, examples and vision.”

Translated from Dutch

Summary of the review: Peter says Thierry's enthusiasm inspired him to start using AI that very evening.

Peter StokCreativeHetMarketingTeam

How to work with me

  • Adoption scan

    Everyone responds anonymously to five short statements. You see where each team is getting stuck and what the first step should be. After six to eight weeks, we measure again.

  • Exploration

    Within a few weeks, a list of opportunities with owners, and a plan for the order in which to tackle them.

  • Temporarily in your team

    usually six months, three days a week. We build, test and hand over.

Book an adoption assessmentWant to start yourself? Explore the AI products

Let's get started

Where is your team getting stuck?

The adoption scan shows you for each team.

Book an adoption assessment