Proof

Four references,at different stages.

Each case shows the entry point, what we did and what changed. We only list results we can evidence. Client names are withheld.

Where each reference stands
  1. Delivered

    Global life sciences company

    From underused licenses to an enterprise fluency program

    Full program with hub, agent and cockpit in production.

  2. Delivering

    US member-owned cooperative

    An executive program as the entry point

    Four executive sessions with individual pre-meets, in delivery now.

  3. Delivering

    Electric and gas utility

    Workforce fluency that becomes the enablement layer for every AI program

    A front door, a learning agent and hackathons, now extending to the client's other AI programs.

  4. Pursuit, in shaping

    Global energy technology company

    Connecting fluency to an agentic process

    Scope being agreed. No outcomes to report yet.

Reference 1 · Global life sciences company

From underused licenses to an enterprise fluency program

Entry point: Integrated program with the full product suite

Delivered

Situation

The company had invested in Copilot, but usage was uneven. Leadership wanted a way to turn the licenses into everyday capability and to see what was changing.

What we did

WhatWhat people learned

  • AI fundamentals for every employee
  • Persona-based advanced learning paths
  • Learning-by-doing playbooks on real work tasks

WhoWho made it stick

  • Executive sessions, shifted toward one-to-one discovery and tailored upskilling
  • A manager toolkit for coaching teams on AI
  • Early adopters and champions, with train-the-trainer material

HowHow it keeps going

  • Engagement Hub, Learning Agent and Value & Adoption Cockpit, deployed in stages
  • Challenges, leaderboards and gamification
  • Badges and a certification framework
  • A communications toolkit

What changed

  • ~15,000employees in scope for foundational learning
  • 150+subject matter experts engaged, including the full executive team
  • 100early adopters in advanced Copilot training
  • 4,033learners assigned AI fundamentals, to complete within three months
  • 1,781learners completed level 1 in month 2
  • 1,018learners completed level 2 in month 2
  • 74Net Promoter Score for the learning
  • Hub, Agent and Cockpit live in production
  • Measurement covers literacy, adoption and estimated productivity. Attribution to business outcomes is still in progress.
  • The entire C-suite trained in AI fundamentals

What we learned

  • Leaders got more from one-to-one work on their own agenda than from group sessions.
  • Persona analysis took more effort than scoped. It needs a named owner and an explicit depth.
  • The hub, the agent and the cockpit are real engineering work, with their own testing, deployment and documentation.
Reference 2 · US member-owned cooperative

An executive program as the entry point

Entry point: Executive-led entry

Delivering

Situation

The leadership team wanted a shared understanding of where AI creates value for members before deciding on a broader rollout.

What we did

Before

Individual pre-meets

One-to-one conversations with each leader on personal AI use, team norms, culture and risks. The four executive sessions were shaped around what we heard.

  1. 1Session 1Use cases and valueAI Foundations
  2. 2Session 2ArchitectureBuilding the Blocks
  3. 3Session 3DataMaking your data ready for AI
  4. 4Session 4Human-AI chemistryHuman-AI Chemistry

What comes out of it

  • A roadmap the leaders keep

    The final session ties AI to the cooperative's purpose and member value: what to do now, what is new and what comes next.

  • A packaged methodology we reuse

    The sessions and the pre-meet approach, ready for the next leadership team.

What changed

  • The executive program is being delivered now.
  • Continuation formats and a broader workforce program are being shaped. Neither is in delivery yet.

What we learned

  • Pre-meets let us build each session around the leaders' own agenda and language.
  • A clear executive entry point gives the organization a low-risk first step before an enterprise program.
Reference 3 · Electric and gas utility

Workforce fluency that becomes the enablement layer for every AI program

Entry point: Workforce fluency program

Delivering

Situation

AI and data tools were spread across business units, and engagement varied widely. Each AI program ran its own enablement. The client wanted one coherent way for employees to find, learn and use AI.

What we did

  • AI program 1
  • AI program 2
  • AI program 3
  • AI program 4

The client's other AI programs

One enablement layer for every AI program

  • Front doorHow

    One entry point for AI tools, training and support

  • Learning agentHow

    Routes each employee by persona and fluency stage

  • Persona pathwaysWhat

    Role-based learning built from persona research

  • HackathonsWhat

    Hands-on agent builds with technology partners

  • Champion networkWho

    Peer champions now, with a manager toolkit next

  • Value cockpitHow

    Literacy, adoption and value tracked in one view

What changed

  • 424employees trained across five hands-on sessions, May to September 2026
  • 127hackathon participants across four events
  • 18working agents built in the hackathons
  • 415employees used the learning passport at an AI expo
  • 80%of surveyed hackathon participants felt more confident using AI
  • The front door and a shared KPI framework are being extended to four other AI programs
  • The front door, the learning agent, persona pathways and hackathons are running.
  • The client is extending the front door and a shared KPI framework to its other AI programs. This is in progress.

What we learned

  • Run fluency as one layer across AI programs, so each new launch lands on a trained workforce with shared measurement.
  • Enablement was siloed per program at the start. Joining it up takes deliberate work and a shared KPI framework.
Reference 4 · Global energy technology company

Connecting fluency to an agentic process

Entry point: Agentic use-case track

Pursuit, in shaping

Situation

The company is building AI agents into a specific business process. The people who run that process need to understand how their work and decisions will change.

What we did

  • Scope in shaping: a change impact assessment for the selected use case
  • A change strategy and a high-level human-AI operating model
  • Recommendations for deployment and scaling

What changed

  • Scope is still being agreed. There are no outcomes to report yet.

What we learned

  • Fluency and agentic process redesign are separate scope decisions, and clients may buy one without the other.
  • Production deployment, learning-content creation and change execution need to be named explicitly as in or out of scope.
A broader track record

Related AI adoption and enablement programsby Capgemini.

  • Energy

    Generative AI and Copilot upskilling for a global workforce

    30,000 employees trained, in 118 countries and 7 languages

  • Automotive

    AI fluency and change leadership for every leadership tier

    Designed to reach about 17,000 leaders in Europe, Asia-Pacific and North America

  • Life sciences

    Gemini adoption with a KPI-driven campaign and knowledge hubs

    90,000 employees in 150 countries, with 7,800 active community members

  • Pharmaceuticals

    Microsoft 365 Copilot rollout with use cases, dashboards and champions

    48,000 employees onboarded, with a champions network of more than 600

  • Aerospace

    New ways of working and Gemini upskilling, including frontline workers

    More than 8,000 long-term champions, in a program reaching 112,000 employees

  • Technology services

    A GenAI campus for a global workforce

    Learning paths by tool, skill and role, with monthly global and local reporting on adoption

  • Transportation

    A prompt academy for a common generative AI culture

    A training academy, a prompt community, expert channels and a community of trainers