AI training for teams on your own tasks, with usage policy and prompt library
We train departments to use ChatGPT, Claude and Gemini on their daily work. Not a generic lecture, but practice on your documents, emails and reports. After the programme the team keeps a usage policy and a shared prompt library.
What hurts
Staff use ChatGPT quietly, from personal accounts. Some paste client contracts and CRM data into it. Security either does not know or bans everything at once.
Two or three enthusiasts get real value, the rest tried once and gave up. There are no shared methods and everyone invents their own. Results depend on the person, not the process.
Online courses teach generic skills. Your accountant, lawyer and sales manager need different things. Nobody has shown them how to speed up their own work.
What we build
- Survey of departments and their typical tasks before training
- Workshops per department on real team tasks
- Basics: how models work and where they fail
- Data safety rules aligned with GDPR: what can and cannot be shared
- Prompt library for each department’s tasks
- Internal AI usage policy
- Practical assignments with feedback
- Follow-up Q&A session two weeks after training
What you get
- Workshop recordings and onboarding materials for new staff
- Prompt library in Google Docs, Microsoft 365 or your knowledge base
- AI usage policy ready for sign off
- List of tasks worth automating next
- Attendance and assignment results report
How we measure the result
- Share of staff using AI every week
- Self-reported time on typical tasks before and after
- Prompts added to the library by the team
- Data policy violations after training
Systems and stack
Industries
Questions
How much does AI training cost?
From $5,000 for a 2–4 week programme. The price depends on the number of departments and groups and how deeply we tailor content to your tasks. One department fits into two weeks. Several departments with their own libraries need longer.
Is training online or on-site?
Both. Online suits distributed teams across Poland and the EU. On-site training works better for hands-on sessions with managers. We record sessions in both formats so new staff can go through them later.
Which model do you recommend: ChatGPT, Claude or Gemini?
It depends on your tasks and security rules. We show all three and compare them on your examples. For confidential data we recommend business plans with no training on your data, or a private deployment.
Will you help with a policy for security and the DPO?
Yes, the policy is part of the programme. It covers what data may be entered, which tools are approved and how to check AI output. Since we are ISO 27001 certified ourselves, we write it in terms your security team accepts. It also notes AI literacy duties under the EU AI Act.
What happens after training?
Two weeks later we hold a Q&A on real cases from the team. Often the team then brings tasks worth automating. These can go into an AI audit or straight into a discovery sprint.
Related services
More in this section
Describe the task. We reply within one business day, the first call is with the founder.
Describe the task. The first call is with the founder.
We reply within one business day. 30 minutes to tell you whether it is worth doing and what it costs.