Participants do not leave with slides about agents. They leave with a first working agent, a clear use case, and rules for using it safely.
- Length: 1 day (8 hours)
- Price: from 70,000 CZK ex-VAT per group, or from 18,000 CZK / participant for open seats
- Format: on-site, custom-built for your team, on demand
- Max participants: 15
- Who it's for: operations, sales, marketing, HR, finance, support, PMs, and process owners
- Prerequisites: one recurring process or scenario people already handle by hand
- A first AI agent for a concrete internal scenario.
- A decision model for when a prompt, skill, or agent is the right tool.
- A map of inputs and outputs the agent needs to work reliably.
- Rules for human review, sensitive data, and handoff of results.
- A backlog of next automations worth solving after the workshop.
Morning: choose the right scenario
Block 1: What an agent means in practice
- Difference between a chat, template, skill, and agent.
- Where an agent saves work and where it only adds risk.
- Examples from internal operations: prep work, document checks, summaries, reports, and drafted replies.
- Participants describe one real process.
- We mark inputs, decisions, outputs, sensitive data, and points where a human must decide.
- We choose a scenario small enough for a first build and useful enough for the team.
Afternoon: agent and operating rules
Block 3: Build the first agent
- We prepare instructions, context, and output format.
- We test the agent on real or anonymized examples.
- We improve it based on concrete failures, not first impressions.
- What the agent may and may not receive as input.
- When the output is a draft, supporting material, or a decision.
- How to collect feedback and improve instructions over time.