AI in a medical facility

Key information

During the training, participants will find out, among other things:

• How to build effective prompts with the KREDO method
• How to use ChatGPT in office work and in LibreOffice Calc spreadsheets
• How to verify AI answers, especially in legal matters
• How to create and edit graphics with the help of AI
• What data should not be shared with public AI tools and what follows from the AI Act
• How to delegate tasks to the ChatGPT Work agent
• How to create skills and use plugins in ChatGPT
• How to work with Gemini and Gemini Notebook
• How to recognise fake websites, cloned voices and impersonation attempts
• How to choose an AI tool for an organisation

The first two days are devoted to ChatGPT, the third day to Gemini and an overview of other solutions: Claude Cowork, Microsoft Copilot and Gamma. The exercises are based on scenarios prepared by the trainer.

Objective

The objective of the training is to familiarise participants with the use of generative AI tools in the office and management work of a medical facility. The training covers work with ChatGPT, including the ChatGPT Work agent, skills and plugins, as well as Gemini and Gemini Notebook. Particular emphasis is placed on patient data security, the legal framework arising from GDPR and the AI Act, and recognising the threats associated with the use of AI for impersonation and manipulation.

Learning outcomes

After completing the training, participants:

• They build prompts with the KREDO method and verify AI answers before using them
• They use ChatGPT for drafting documents and working with spreadsheets
• They distinguish between data that can and cannot be shared with public AI tools
• They delegate tasks to the ChatGPT Work agent and create their own skills
• They work with Gemini and Gemini Notebook on uploaded documents
• They recognise typical AI-enabled impersonation attempts

Programme

The programme includes theoretical modules and practical exercises.

Day 1: ChatGPT, graphics, data security and the AI Act

Module 1: The basics of generative AI

• Sharing participants’ experience
• Key concepts related to how AI works (prompt, token, hallucination, context)
• How to act in the face of false content created by AI
• Comparison of the key AI models

Module 2: Prompt engineering: the KREDO method

• The anatomy of an effective prompt: the KREDO model
• The most useful prompts to start with
• Building KREDO prompts for office and management scenarios
• Fast vs advanced models: when to use which

Module 3: ChatGPT in office and management work

• Drafting letters, notes and summaries
• Working with LibreOffice Calc spreadsheets with the help of ChatGPT
• Legal and organisational questions, including those concerning cooperation with the NFZ (National Health Fund)
• Deep Research: searching for information with sources
• Verifying AI answers before using them

Module 4: Creating graphics in ChatGPT

• Building an image prompt
• Editing an image in successive steps
• Information materials for patients
• Retouching and reworking photos
• Image and copyright in AI-generated graphics

Module 5: Introduction to skills in ChatGPT

• What skills are and how they differ from an ordinary prompt
• When it is worth saving repeatable work as a skill
• Demonstration: a skill that prepares a document according to a template

Module 6: Medical data, GDPR and the AI Act

• Health data as a special category of personal data
• Where data entered into ChatGPT goes
• What may and may not be shared with public AI tools
• Anonymisation and pseudonymisation
• Shadow AI: personal accounts used for work tasks
• The AI Act in healthcare
• ChatGPT account privacy settings

Day 2: ChatGPT Work, skills and plugins

Module 7: ChatGPT Work: working with files

• What ChatGPT Work is
• Working on local LibreOffice files
• Documents prepared according to the facility’s template
• Local work versus work in the cloud

Module 8: ChatGPT Work: automation and the browser

• Collecting information from websites in the built-in browser
• Scheduled tasks
• Pages and dashboards created by the agent
• Agent permissions and prompt injection

Module 9: Skills in ChatGPT

• Building your own skill
• Testing and improving skills
• A skill that prepares a summary from a spreadsheet
• Updating a skill based on daily work

Module 10: Plugins in ChatGPT

• What plugins are and how to install them
• Connecting ChatGPT to Gmail and Google Drive
• Plugin permissions and patient data
• Two-day summary and Q&A session

Day 3: Gemini, AI and cybersecurity, overview of other tools

Module 11: Gemini

• Comparing Gemini and ChatGPT on the same tasks
• Skills in Gemini
• Deep Research and Canvas
• Gemini in Gmail and Google Drive

Module 12: Gemini Notebook: working with sources uploaded to the tool

• What Gemini Notebook is and when it works better than a chat
• Questions about uploaded regulations and procedures with references to sources
• Audio summaries, mind maps and quizzes
• The facility’s knowledge base

Module 13: Two sides of technology: AI and cybersecurity

• Building a fake website in Gemini
• Voice cloning: demonstration
• A voicebot on the helpline: demonstration
• Verifying the identity of the person on the other end at reception

Module 14: Claude Cowork: presentation

• The Claude ecosystem
• Claude Cowork as an agent working on files
• Comparison with ChatGPT Work

Module 15: Microsoft Copilot and Gamma: presentation

• Copilot and Work IQ in Microsoft Office applications
• Gamma: information materials and documents

Module 16: Choosing an AI tool and summary

• How to choose an AI tool for an organisation
• Personal and business plans
• Training summary and Q&A session

Trainer

Franciszek Kołpanowicz: entrepreneur and certified AI trainer who helps leaders and teams boost their effectiveness with intelligent technologies. He has been published in Rzeczpospolita, among others. He has over 2,000 hours of experience working with generative AI and shares his knowledge at public events, including the “Manager in the Age of AI” training organised by Rzeczpospolita, as well as at the UKEN and PASE conferences.

His background in philosophy (Jagiellonian University) and his experience in IT quality control allow him to combine analytical thinking with asking the right questions. His own practical journey with AI showed him how this technology can become a real lever for growth.

Drawing on hands-on experience, he shows leaders and their teams how to use AI not only to work more efficiently but also to think more clearly and make faster business decisions. His AI training has, among other results, cut the time needed for routine analysis of multi-page legal documentation by 80%.

In his training sessions he pays particular attention to the participants themselves, drawing on their experience and needs as the work progresses. As a result, his sessions are not repetitive routine but an enriching and individual experience.

His mission is to make artificial intelligence a practical tool for development: business, intellectual and cultural.

Prerequisites

Basic computer and web browser skills. The training does not require any previous experience with AI tools.

Technical requirements for participants:

• A ChatGPT Plus account, at least one per three workstations
• The ChatGPT desktop app (Windows or Mac) installed before the second day of the training
• A Google account logged in on the computer for the third day of the training

Organisational requirements on the client’s side

• A computer for every participant doing the exercises, with no sharing of workstations
• A camera and a microphone or headphones
• Access to the Zoom platform
• A stable internet connection
• The ability to install applications on participants’ computers

The first two days work with a single tool only, ChatGPT, so that participants have time to genuinely master it in the context of their own work, rather than just seeing a few demos. The third day is devoted to Gemini and an overview of other solutions.

Key information

Training for doctors who want to use generative artificial intelligence in the part of their work that happens outside the patient examination: in documentation and correspondence, preparing information materials, working with the literature, reporting and teaching presentations. The session takes the form of a lecture with live demonstrations of tools, based on examples prepared by the trainer.

During the training, participants will find out, among other things:

• How language models work and how Microsoft Copilot, Gemini and ChatGPT differ from one another
• How to recognise a model’s hallucination and bias before its answer ends up in a document
• How to write prompts with the KREDO method to get useful results the first time
• What data may be shared with a public AI tool and what must not be taken outside the healthcare provider
• How to use AI in documentation, correspondence and communication with patients
• How to work with guidelines and literature with a limited risk of hallucination

The training includes an overview of the three tools most commonly found in organisations, namely Microsoft Copilot, Gemini and ChatGPT, together with their strengths and weaknesses and selection criteria. Tools that answer solely on the basis of uploaded documents, such as NotebookLM and Copilot Notebooks, are covered separately; they are useful wherever an answer must be grounded in a specified source.

Objective

The objective of the training is to familiarise participants with the capabilities of generative artificial intelligence in a doctor’s work, covering an overview of Microsoft Copilot, Gemini and ChatGPT, prompting with the KREDO method, applications in documentation, correspondence and communication with patients, and work with the literature. The training discusses the risks associated with model hallucinations and bias, as well as the rules for handling health data and medical confidentiality in public AI tools. The emphasis is on critically evaluating results and making a conscious choice of tool for a specific task.

Programme

The programme covers 3 hours in lecture format, with live tool demonstrations and room for questions from the audience.

Module 1: The basics of generative AI

• Key concepts related to how AI works (prompt, token, hallucination, context)
• Capabilities of language models: working with text, documentation and data
• Fast models versus models with advanced reasoning: when to reach for which
• Comparison of the key AI models: Microsoft Copilot, Gemini and ChatGPT

Module 2: Risks of language models and the user’s responsibility

• Hallucinations: where they come from and how to act when faced with false AI-generated content
• Bias and sycophancy of the model, shown with two opposing prompts
• Shadow AI: personal accounts used for work tasks and the risk of data leaks
• The four competencies of an AI user in the 4D model: delegation, description, discernment, diligence

Module 3: Patient data, medical confidentiality and the legal framework

• Health data as a special category of personal data
• Medical confidentiality and sharing content with public AI tools
• Anonymisation and pseudonymisation before entering content into a tool
• The AI Act and the classification of artificial intelligence uses in healthcare
• Account privacy settings and control over conversation history

Module 4: Applications in a doctor’s daily work

• The KREDO method: a way to improve the quality of AI answers
• The most useful commands to start with
• Documentation and correspondence: opinions, certificates, appeals, applications, reports
• Communication with patients: translating medical language, information materials and recommendations
• Working with guidelines and literature in tools that answer on the basis of uploaded documents
• Teaching and presentations: slide decks, lectures, materials for students and residents
• Working with data and summaries: schedules, reporting, spreadsheet analysis
• Translation and working with foreign-language text

Module 5: The Microsoft 365 ecosystem and additional tools

• Copilot in Outlook, Word, Excel, PowerPoint and Teams
• Meeting summaries and extracting decisions and tasks from transcripts
• Tools that answer solely on the basis of uploaded documents: NotebookLM and Copilot Notebooks
• Five questions that help you choose an AI tool for an organisation
• Summary and Q&A session

Trainer

Franciszek Kołpanowicz, entrepreneur and certified AI trainer. He has been published in Rzeczpospolita, among others (an article on NotebookLM as a company’s digital brain, November 2025). He has shared his knowledge at public events: at the training “Manager in the Age of AI” organised by Rzeczpospolita, at the UKEN, PASE and ICBU conferences, and at the seminar “AI, AR and Autonomous Robotisation 2026”.

He graduated in philosophy from the Jagiellonian University and spent years working in IT quality assurance. This combination of analytical thinking, the philosophical habit of asking precise questions and a technical understanding of systems allows him to explain how AI works in the language of everyday work. He is a certified trainer (Trainers’ School, Jagiellonian University Extension School). He has over 2,300 hours of work with generative AI and over 2,000 hours of working with groups.

He has delivered training for public institutions, cultural institutions, and manufacturing, logistics and retail companies, and has also prepared internal AI training materials for a nationwide education network. In his training he pays particular attention to the participants and their real needs, and chooses demonstration scenarios to match the tasks the group performs.

Prerequisites

The training does not require any previous experience with generative AI tools or technical preparation on the participants’ side. The lecture format does not involve participants working on their own equipment, so no accounts or licences in AI tools are needed to take part.

Organisational requirements on the client’s side

• A lecture room for the stated number of participants, with a clear view of the screen from every seat
• A projector or large screen with a resolution that allows the tool interface to be read from the back row
• A sound system and microphone for the trainer
• A wireless microphone or staff to pass the microphone for questions from the audience
• A stable internet connection available at the trainer’s workstation, necessary for live demonstrations
• A power socket at the trainer’s workstation and an HDMI connection or adapter

The training takes the form of a lecture with demonstrations, so participants do not need laptops or prepared accounts in AI tools.

To be confirmed before the training

Ecosystem and tools:

• Which office suite does the organisation use: Microsoft 365 or Google Workspace? This determines which ecosystem we base the demonstrations on.
• Does the organisation provide doctors with a specific AI tool, for example Microsoft 365 Copilot on a work licence?
• Does the organisation have a policy on the use of AI tools that the training should refer to?

Infrastructure:

• A stable internet connection for the trainer, also available to a guest from outside the organisation