// services.ai_training Module 05 of 06

AI Training
for your team.

Practical adoption of Claude, Copilot, ChatGPT and Google Gemini in your team's real work. No theory - with ready-made ways of working you put in place straight away.

haski.io / training.run
$ train --team
[ consultation ]45 min · free
[ program ]tailored
[ follow-up ]included
[ tools ]claude · copilot · gpt · bielik
Initial consultation
Free
Tools in the stack
4platforms
Program
tailored
// problem >> solution

Problems we solve

training.schema
2 tables · 6 fields
current_state
tried_gave_up The team typed a few questions into ChatGPT, got a generic answer, and gave up. No proven way to talk to AI, no concrete examples for a given role - so "AI doesn't work for us".
drop-off
everyone_uses_different The accountant uses free ChatGPT, the lawyer Claude, the director Copilot. No standard, no shared prompts for AI, no control over which client data ends up in AI. Knowledge doesn't compound.
silos
fear_for_the_client You know AI could help, but clients have signed confidentiality agreements and GDPR is always at the back of your mind. Without a policy, without tools that guarantee data stays in the EU - you'd rather the team didn't use it at all.
GDPR
target_state
framework_per_department Every person on the team walks away with ready-made ways of working where AI genuinely saves time. The example is taken from the participant's actual work, not a demo. We set the goal and the measure of success together on the initial call.
workflow
one_stack The team gets a recommendation: when to use Claude (long analyses, Polish text), when Copilot (Microsoft 365), when ChatGPT, when Bielik on your own server.
standard
ai_policy_gdpr A template AI policy for the company, clauses for client contracts, a list of sensitive data mapped to tools. The Polish Bielik model on your own server for the most confidential data - with nothing sent outside.
compliance
// pain.delta fk → impact
// scope

What you get from AI training

AI adoption from scratch · level 0 → 3

What text AI is, where it gets things wrong, when not to use it. A skills test at the start and at the end. We start from examples, not from the history of AI.

Prompting framework

A "role + background + task + format + constraints" template adapted to your industry. A framework for writing skills so the team can keep improving on its own.

Use cases in every department

Accounting (cost allocation, drafting replies to the tax office), law (contract review, clauses), finance (profit-and-loss analysis, forecasts).

Choosing the tools

Deciding when to use Claude, when Copilot, when ChatGPT, when Bielik on your own server - depending on the task, data sensitivity and the licences you already have.

AI policy + GDPR

A template AI policy document for the company.

Post-training support

Post-training support over a period matched to the team: a Slack/Teams channel for questions, a live Q&A session, a review of selected participant tasks. No "the workshop's over, you're on your own".

// stack
[ legal / analytical ]
Claude
[ M365 native ]
Microsoft Copilot
[ GPT ecosystem ]
ChatGPT Enterprise
[ local · PL ]
Bielik 11B
// faq

Frequently asked questions

Does training make sense if the team doesn't use AI yet?

Yes - it is actually the better moment. We start from level 0 (what text AI is, what it cannot do, where it gets things wrong), then concrete use cases for every department and ready-made ways of working you can put in place right away.

What about GDPR and sensitive client data?

We pick tools and licences that meet the highest data-protection standards - compliant with EU law, with a guarantee that data is neither stored nor used to train AI, and for the most sensitive data, AI running on your own server with nothing sent outside. We also help you write an AI policy for your company.

What results can we expect after the training?

We discuss results individually for each specific team. Client A wants to learn how to feed AI the right information so it answers accurately. Client B wants to build their own reusable templates and prompts. We choose the goal and a measurable result together on the initial call, before we sign off on the scope.

How long is the training and what is the format?

Also decided individually. We tailor the format, intensity and duration to the size of the team, the starting level and the goal of the training. You get a concrete proposal after the initial call.

// next_step No commitment

Adopt AI
the smart way

First we get to know the team, point out 2-3 of the most frequent tasks to shorten, and show which training format fits your first group.