Why switching on Copilot isn’t enough.
A licence makes AI available, but not yet effective. That takes a platform that fits the organisation, clear rules that give confidence, and people who know how to use AI for their work.
Creating fields of gravity
- Key users: trained, connected, anchored in their own department
- Colleagues who get started through them
- A department that follows in the next wave
Where lasting adoption fails in everyday work
AI is usually in the company long before there is a platform. And where a platform has been bought, adoption often ends after the kick-off.
You’re in the right place if AI excites individuals in your company but doesn’t reach the organisation.
Not yet clear which use cases pay off? Then Getting started with AI is the better starting point. If your case fits no standard tool, it’s a development project.
- Not approvedShadow AIPrivate accounts, customer data leaves the company
- Barely usedPlatform licencesAssigned, kick-off has taken place
- UnresolvedUsage policyQuestions from data protection, works council, IT security
- OpenOwnership of the rolloutNo key users, no route for new use cases
What we take on in AI adoption
We introduce an AI platform in your company, from selection to everyday use. Each building block can also be commissioned on its own.
- Platform selection
- We capture your requirements independently of any platform and recommend the solution that fits them best.
- See the criteria
- Technical integration
- We set up the platform with your IT: single sign-on, roles and permissions, connection of your knowledge sources, security tests.
- Governance
- We work out the usage policy, the role model and the approval process for new use cases with you. We involve data protection, the works council and IT security and provide the basis for a works agreement.
- What the EU AI Act requires
- Pilot and rollout
- We implement the first use cases with your departments and roll the platform out across the company in waves.
- See the approach
- Enablement
- We train key users and staff in impulse sessions, workshops, leadership trainings and hackathons. The courses run on the Nao Training Hub.
- To the Training Hub
Which platform is the right one?
Microsoft Copilot, Langdock or Claude Enterprise: we recommend the platform that fits your requirements. We check four things.
- Knowledge
- The platform accesses your documents: SharePoint, document storage, QM system.
- Models
- You use the language models that fit your tasks and can switch between them.
- Departments
- Employees create their own assistants without IT building each one.
- IT environment
- Data stays in the EU and is not used for training. Sign-in and permissions come from your IT.
What a platform has to deliver
Five phases from kick-off to productive use
The first use cases are productive after six to eight weeks, the company-wide rollout is complete after about five months. Technical integration runs in parallel with the use-case survey.
Setup
Kick-off, roles and responsibilities, survey of use cases with departments and IT. Assessment by value, feasibility and data situation, selection of the pilot cases.
Integration
Platform provisioning, single sign-on, connection of your knowledge sources, roles and permissions, integration and security tests with your IT.
Pilot
Delivery of the prioritised use cases in small teams from the department and Nao, in one- to two-week cycles. Training of the key users.
Rollout
Introduction in waves across the whole organisation, supported by communication and onboarding. In parallel, usage policy, role model and the basis for a works agreement take shape.
Operation
User support, broad training, further use cases. From here you decide how much support you still need.
How your people become AI-ready
What happens with the people in every wave: adoption doesn’t come from one training, but from a sequence. Four stages that start in the pilot and repeat in every rollout wave.
Show & Tell
We present the platform and explain how generative AI works, how to write good prompts and what safe use means. In impulse sessions of 60 to 90 minutes, with explainer videos and quick-start guides.
See & Try
Users work with it themselves for the first time: on prepared tasks and on their own, well-defined cases. Small-group workshops plus office hours.
Collaborate & Learn
Use cases, assistants and a prompt library take shape in co-creation workshops with the departments.
Lead & Multiply
The key users carry the topic on: train-the-trainer, regular key-user rounds, handover of the materials. In addition, when pace and visibility are needed: leadership training and hackathons.
A vetted platform for a whole organisation
Interest in AI among staff was high, many already used freely available tools. What was missing was a central, vetted platform with clear security and data protection standards. The public-sector organisation is now introducing it in a structured way: use cases prioritised, Langdock securely connected to the existing document base, pilot with trained key users, rollout and governance under way.
- Platform
- Langdock, RAG to the document base, roles and permissions, SSO
- Approach
- Use cases, pilot with key users, rollout, governance
- Status
- ongoing, in delivery
Does this fit your organisation?
That’s what we work out in a first call: where you stand, which platform is an option and whether our approach fits your organisation.