AI-Powered Workflows

Turn individual AI experiments into reliable workflows: In this workshop, you will learn how to integrate Generative AI into recurring tasks so that good results are no longer a matter of chance.

Many recurring tasks can be effectively supported by Generative AI. However, these efforts often remain one-off experiments because a clear workflow, a reliable prompt, or the right contextual information is missing.

In this workshop, you will develop a structured, reusable prompt for a task from your own work. Between the two online sessions, you will test it in your day-to-day work and refine it based on the results. You can then apply the same approach to other tasks. No programming skills are required.

Course Content

Identify and select suitable tasks for using Generative AI
Describe tasks in a structured way
Create reusable prompts with the relevant contextual information
Test, evaluate, and improve results
Apply the approach to other tasks

Course Structure

Please note: Each workshop consists of two online sessions.

Preparation phase, seven working days
You will identify suitable tasks based on a set of criteria and select two potential tasks to work on.
First online session, 150 minutes
You will select a suitable recurring task from your day-to-day work and map out the associated workflow. You will then develop a reusable prompt that incorporates the necessary contextual information.
Transfer phase, five working days
You will test your prompt in your day-to-day work, document the results, and make initial improvements.
Second online session, 120 minutes
You will reflect on your experience during the transfer phase, refine your prompt, and test the revised version. Finally, you will apply the approach to other tasks from your day-to-day work.

What You Will Take Away

You will leave the workshop with a tested, reusable prompt for one of your own tasks, as well as a transferable approach that you can apply to other use cases.

Target Audience

This workshop is designed for employees working in administration and management across the Helmholtz Association who want to use Generative AI more systematically for recurring tasks.

The workshop will be held in German.

Requirements

This course is intended for participants who already have some practical experience with Generative AI. It requires basic knowledge of working with Generative AI, such as that covered in the course “Using ChatGPT & Co. Effectively.”

To participate, you will need access to an AI tool such as ChatGPT, Copilot, Claude, or Blablador, a suitable task from your day-to-day work, and sufficient time to test your prompt between the two online sessions.

To help us prepare the workshop, you will be asked to submit two potential tasks using a short template before the first session. If you do not have suitable access to an AI tool, HIDA can provide temporary access for the duration of the workshop.

Course Dates

Please note: Each workshop consists of two sessions.

27. October 2026, 09:00 - 11:30 am & 3. November 2026, 09:00 - 11:00 am
10. November 2026, 09:00 - 11:30 am & 17. November 2026, 09:00 - 11:00 am

Places are limited to 25 participants and will be allocated on a first-come, first-served basis.

The Trainer

Jonas Kuppler

Jonas Kuppler is a science manager at the Geo.X research network at the GFZ Helmholtz Centre for Geosciences in Potsdam. There, he designs and coordinates interdisciplinary training and networking programs for researchers.

A particular focus of his work is on learning formats related to generative AI and data science skills, including the Geo.X Young AI Academy. In his spare time, he occasionally still writes scientific publications about bees and flowers.