Vielen Dank für die Zusendung Ihrer Anfrage! Eines unserer Teammitglieder wird Sie in Kürze kontaktieren.
Vielen Dank, dass Sie Ihre Buchung abgeschickt haben! Eines unserer Teammitglieder wird Sie in Kürze kontaktieren.
Dauer 7 Stunden
Schulungsübersicht
- Block 1: Foundation and techniques
- Introduction and assessment
- Live survey on the current level of knowledge
- Tasks brought by the group
- What happens in the model when prompting, context window and invented details
- Exercise: Each participant presents an own task to work on during the workshop.
- Building blocks of a good prompt
- Role, task, context, examples, format and limits
- Weak and strong prompt in a live comparison
- Specifying audience, tone and one or two good examples
- Language and language level, for example plain language or B1
- Exercise: Improve a prompt for a real task of the group in three rounds.
- Breaking down and improving tasks
- Creating a plan first, then letting it run
- Using reasoning models correctly
- Allowing clarifying questions and finding errors
- Templates with placeholders and output as a table
- Exercise: Break down a multi-step task into steps and save it as a template.
- Outcome: Each participant has at least one verified template for their own work.
- Introduction and assessment
- Block 2: Working lastingly and safely, practical workshop
- Prompting on a lasting basis
- Levels of chat, instructions, memory, project, skill and agent
- Personal instructions that hold
- Projects for the team with a style guide or reply templates
- Own assistants as a skill, custom GPT or Gem
- Exercise: Create a shared project with a standing instruction for a standard task of the group.
- Working with tools
- Querying and comparing Excel, PDF and Word
- Having scans and screenshots read
- Formulating research assignments and checking sources
- Dictating instead of typing
- Exercise: Have a table or a document from the own area analyzed.
- Working safely
- Recognizing invented details and requesting supporting passages
- Having figures calculated by a formula or the analysis tool
- What may go into which model, anonymizing cases
- Hidden instructions in emails and PDFs
- Exercise: Cross-check an AI result in a targeted way and document the errors.
- Practical workshop
- Group rooms with two to three real tasks from the area
- The trainer supports the rooms
- Creating the group's prompt collection
- Exercise: Work on two tasks of the own group through to the finished result in the workshop.
- Outcome: Each group has a shared prompt collection, a shared project and two solved tasks from real daily work.
- Prompting on a lasting basis
Voraussetzungen
- Access to the AI tool approved in the company (Claude, ChatGPT, Copilot or Gemini), up to five use cases per group in advance, anonymized sample documents.
- No programming skills required.
Audience
- Teams from case processing, specialist departments, research or IT support who already have first experience with AI and want better results for their own tasks.
Erfahrungsberichte (3)
Umfassende Informationen zum Hintergrund; wie man eine Prompt-Anfrage formuliert.
Olaf Kannt - Helios Klinikum Krefeld
Kurs - Generative AI and Prompt Engineering in Healthcare
Maschinelle Übersetzung
die Tipps und empfohlenen Prompts, die wir aus diesem Training mitnehmen können
Lee Mei Lin - ST Engineering IHQ Pte Ltd
Kurs - InVideo AI: Creating Engaging Short-Form Videos
Maschinelle Übersetzung
Einsatz von angemessenen und wirksamen Prompts
Marses Pacaldo
Kurs - Generative AI and Prompt Engineering for Corporate Professionals
Maschinelle Übersetzung
