[ FOR AGENCIES & IN-HOUSE TEAMS ]
One capability partnership. AI capability, creative craft and literacy for European agencies and in-house brand teams.
[ WHY NOW ]
01 · The law
The EU AI Act's Article 4 has applied since 2 February 2025, and enforcement began in August 2026.
02 · The bar
It asks for measures proportionate to your role and risk, not one mandated level. You decide the standard. We build it with you.
03 · The gap
Most teams still lack a consistent programme, evidence trail and adoption standard. Build those properly and governance becomes a commercial advantage.
[ THE PARTNERSHIP ]
Up to 25 seats
The whole team on all 12 lectures, each with their own Harness.
Private cohort
A kickoff onboarding and monthly live masterclasses with the faculty.
Team AI OS
A shared Harness tuned to your brand, prompts and workflows.
Two workshop days
On site or remote, per year, built around your team.
Train the Trainer
We certify an internal AI Lead so the capability stays in house.
Team certification
Everyone graded on a real brief, verifiable in the directory.
Quarterly review
Leadership reporting on adoption, output and real change.
Article 4 pack
Documentable evidence of staff literacy and governance readiness.
Named partner
A direct line and a private priority channel.
Marketplace and ARK
Full asset access, and a preferred on-ramp to ARK.
[ HOW IT WORKS ]
We map where the team is today, tool by tool, person by person.
The private cohort starts. Kickoff, seats live, everyone building in week one.
Real briefs, the Harness, graded work. Skill you can point at.
Quarterly with leadership. What moved, what is next, where it scales.
[ CHOOSE THE RIGHT START ]
Start with one priority team, prove adoption on real work, then roll the operating model across the organisation. Every proposal is scoped around seats, delivery, reporting and integration needs.
01 · One-time · 6 to 8 weeks · one priority team
€15,000 to €25,000
A focused diagnostic and delivery sprint that establishes the baseline, ships real workflows and produces a rollout decision.
02 · Recurring · approximately 25 to 100 seats
€60,000 to €90,000 / year
A managed capability programme for a team that needs consistent craft, adoption, evidence and governance across the year.
03 · Recurring · multiple teams, regions or business units
€120,000 to €250,000 / year
A governed operating model for organisations scaling AI capability across functions without fragmenting standards or evidence.
Exact pricing depends on seats, delivery format, reporting depth and integration scope. Annual contracts are recurring revenue; pilots, Academy and Marketplace purchases are one-time revenue.
Need a 90-day transformation partner beyond capability building? That is ARK.
See ARK[ Executive toolkit ]
Nine working tools for the people who set the AI agenda: where to look for opportunity and risk, how to decide, how to prove it, and how to keep leadership informed.
Last reviewed: 9 September 2026
The NIST AI Risk Management Framework organises AI work into four functions: Govern, Map, Measure and Manage. Mapping the context and the use cases comes before measuring or managing anything.
When to use it
At the start of any AI programme, and again each quarter, before you decide where to invest or where to slow down.
A short intake form turns an informal request into something that can be governed. It is the Map function of the NIST AI RMF applied to one request at a time.
When to use it
Whenever a team wants to start using AI for a new task, tool or workflow.
A one page memo forces a clear recommendation instead of an open-ended discussion, and gives leadership a record of what was decided and why.
When to use it
Before committing budget, a vendor or a rollout to more than one team.
The Measure function of the NIST AI RMF asks for tracked, comparable indicators. A scorecard puts every AI initiative on one page so leadership can compare them.
When to use it
Monthly, alongside the monthly leadership report.
Choosing an AI vendor is a governance decision, not just a purchasing one. The checklist below follows the Map and Govern functions of the NIST AI RMF: know what the system does, who is accountable, and what happens when it fails.
When to use it
Before signing with any AI tool or model vendor, and again at each renewal.
The Manage function of the NIST AI RMF calls for a named, resourced response to risk, not a general intention. A human oversight plan states exactly who checks what, and what they do when something looks wrong.
When to use it
For any AI use case that reaches a client, the public, or a regulated decision.
A short, dated roadmap keeps a rollout accountable. It breaks a large change into reviewable phases, in line with how the enterprise pilot to rollout model is scoped.
When to use it
At the start of any team-wide or company-wide AI rollout.
Regular, short reporting is what turns an AI programme from a project into a managed capability. It matches the Measure and Manage functions of the NIST AI RMF applied at the reporting cadence.
When to use it
Every month, or on whatever cadence leadership has agreed.
Article 4 of the EU AI Act asks providers and deployers to take measures ensuring a sufficient level of AI literacy in staff dealing with AI systems. A dated log of who was trained, on what, is the practical record of that effort.
When to use it
Ongoing, updated whenever someone completes training, a lecture, or a certification step.
[ Train the Trainer toolkit ]
Eleven working tools for the internal AI Lead: plan a session, run it well, check what landed, and keep the standard current.
Last reviewed: 9 September 2026
Adult learners bring existing experience and expect training to respect it. A short pre-assessment lets a facilitator meet people where they are instead of teaching to an average that fits no one.
When to use it
Before planning any internal session or onboarding path.
A one page canvas keeps a training session anchored to one goal and one real task, which adult learning guidance identifies as central to engagement and transfer.
When to use it
When planning any internal lab, lecture or workshop.
A minute by minute run of show is what separates a planned session from an improvised one. The CDC Training of Trainers model relies on this kind of structure so a new trainer can deliver consistently.
When to use it
For every session a trainer runs, especially their first few.
Show, then do, then respond, is the core loop behind the CDC Training of Trainers model: a master trainer demonstrates, the new trainer practises, and specific feedback closes the loop.
When to use it
When coaching a new internal trainer, or teaching any hands-on skill.
Asking a learner to teach the material back is one of the clearest tests of understanding used in the Training of Trainers model, since a person cannot teach what they have not actually understood.
When to use it
At the end of a session or a certification step, before marking someone competent.
A shared bank of good questions means every session, not just the best facilitator's session, surfaces the same depth of thinking.
When to use it
Building or refreshing any session, and when a new trainer takes over a topic.
Short, consistent feedback after every session is what lets a trainer improve session by session instead of guessing.
When to use it
At the end of every session, every time, not only when something felt off.
The Training of Trainers model expects the master trainer to keep coaching new trainers after the first session, not just to run one workshop and stop.
When to use it
Every time an experienced trainer coaches a newer one.
A short, frequent pulse check catches drop-off in adoption early, well before it shows up in the monthly leadership report.
When to use it
Every two weeks, across the team being trained.
Learners need to know, in advance, exactly where to take a question the trainer cannot answer, or a result that looks wrong. This mirrors the human oversight and Manage practices in the NIST AI RMF, applied to the training programme itself.
When to use it
Set up once, then shared at the start of every onboarding path or session series.
A certification is only as credible as the evidence behind it. This checklist ties the teach-back, the graded work and the literacy evidence log together into one record per person, in the spirit of Article 4's call for demonstrable AI literacy measures.
When to use it
Before marking anyone certified on the internal standard.
[ SESSION DETAILS ]
Every partnership runs on a fixed rhythm of live sessions. Here is the format, the length and who joins each one, so you can plan calendars before you sign anything.
Session
Format
Length
Who joins
Seats go live, the team meets the faculty and everyone leaves with their first brief already running.
Format
Live, on site or online
Length
Half a day
Who joins
The whole cohort
One craft or capability topic per month, worked on your own material, with questions answered in the session.
Format
Live online, recorded
Length
90 minutes
Who joins
The whole cohort
A real brief taken from blank page to finished output in one day, with the workflow written down before anyone leaves.
Format
On site or remote, twice a year
Length
One full day each
Who joins
The team plus invited stakeholders
Where the field actually stands, what changed since the last briefing and what it means for your decisions this quarter.
Format
Live online, closed room
Length
60 minutes
Who joins
Leadership and board
One person is certified to run the standard internally, so the capability stays with you after the year ends.
Format
Live working sessions
Length
Seven steps across the year
Who joins
Your internal AI Lead
Adoption, graded work and output reviewed against the baseline, with the next quarter agreed in the same meeting.
Format
Live online or on site
Length
60 minutes
Who joins
Leadership and the named partner
Sessions run in English. Russian and Latvian are available on request. Recordings, slides and materials stay in your team space.
[ REQUEST THE OFFER ]
Send the shape of your team and we come back within two working days with a written proposal and a time to talk.
No obligation. Nothing is charged and no seat is reserved until a proposal is signed.
[ QUESTIONS ]
One partner, one number, a team that ships. Tell us the shape of your team and we will send the proposal.
