Claude training: advanced prompting, Projects and Skills for your teams

Who gets the most out of Claude?

Claude is most useful to teams whose work is reading, analysing and producing documents, or applying the same method again and again. It stands out for the quality of its reasoning and writing, and for its ability to carry out long, multi-step tasks.

Its ecosystem is open. Skills (your methods and formats packaged as reusable modules) and MCP connectors (the standard that links Claude to your tools) are open standards: what your teams build stays portable.

This page is about training your teams. For choosing a subscription plan and the rollout steps, read our article Deploying Claude in your company (in French).

What are the use cases by role?

Every role has tasks where Claude genuinely helps, as long as a person validates what matters. Here are examples we work on with teams:

• Legal: spot risky clauses in a long contract, compare two versions of a contract (differences, additions, deletions), always with human validation.
• HR: an assistant that answers internal questions by quoting the exact passage of your agreements; help reading applications, with the decision staying human.
• Sales: a first draft of a proposal ready for review, meeting notes turned into a follow-up and a proposal.
• Marketing: turn a white paper or a study into posts, newsletters and articles, in your tone of voice.
• Leadership: summarise a report into key points and risks, prepare the agenda and talking points for an important meeting.

Claude or another tool: how to choose?

The right tool is the one that fits where your teams work and the data they handle. If they spend their days in Outlook and Teams, also look at Copilot training. You do not always have to pick one: several of our clients allocate licences in tiers, with a core tool for everyone and specialist tools for the teams that need them.

No software vendor pays Datajuba: we train on the tool you have chosen and say plainly what it does well and less well. To decide, we look at five criteria with you:

• your environment: Microsoft 365, Google Workspace or both;
• the data to connect and how sensitive it is;
• your user groups: everyday use, advanced users, agent builders;
• the annual cost per user group, licences and usage included;
• the constraints set by your group and your IT department.

What do you need in place before the training?

Three things: a Claude business account for each participant, clear usage rules and up-to-date reference documents.

• Access: everyone works on their own account. At scoping, a checklist approved with your IT department confirms the plan, the seats available and the connectors allowed.
• The AI Charter: which tools are allowed, which data can be shared with Claude, who signs off. If you do not have one yet, the AI Charter workshop sets it up with your management, IT and HR.
• Your documents: an assistant is only as good as the information it can see. Up-to-date procedures, grids and templates make for better Projects.

What happens after the training?

After the day, everyone keeps working on their own tasks, and the company organises what comes next: champions in each team, agents, then ongoing oversight. Training is the start; leading the change is what makes it last.

• AI Ambassadors: the training reveals the most engaged people. One AI Ambassador per team shares the assistants that work and reports back on needs.
• Agents: an agent carries out a task from start to finish, not just a single answer. Our two-day course on building agents with Claude follows on from this day; our guide Building an AI agent with Claude (in French) gives an overview.
• Oversight: a Chief AI Officer part-time (CAiO) prioritises projects, leads your AI Ambassadors and measures with you the real value of the use cases.