How can Generative AI support data analysis without sharing sensitive original data with external providers? In this workshop, you will develop privacy-friendly workflows using Excel, dummy data, and AI-generated formulas and put them into practice.
Chatbots and AI assistants can support data analysis in various ways, for example by generating formulas, writing code, or helping to structure analytical steps. However, uploading entire sensitive datasets and relying on AI-generated results without verification is often neither appropriate from a data protection perspective nor methodologically sound. Personal or confidential data, in particular, should remain under the user’s control. At the same time, analytical processes need to be transparent, traceable, and reproducible.
This workshop helps you address this challenge with a practical three-step approach. You will break down a specific data analysis task into individual steps, create suitable dummy data, and use an AI assistant to develop Excel formulas and possible solutions. You will also learn how to systematically review AI-generated suggestions and identify errors. By the end of the workshop, you will be able to apply the tested workflow to your local original data in a transparent and reproducible way – without sharing the data with the AI service.
Course Content
- Assess the opportunities and risks of automated data analysis using chatbots
- Break down an analysis task into inputs, desired outcomes, individual steps, and verification steps
- Create dummy data to avoid sharing sensitive original data with external AI services
- Create Excel formulas, tables, and charts with the support of an AI assistant
- Understand AI-generated results, check their plausibility, and apply them locally
Target Audience
This four-hour online course is designed for employees of the Helmholtz Association who analyse data using Excel or a comparable spreadsheet application and already have some basic experience with Generative AI (e.g. through the introductory course “Using ChatGPT & Co. Effectively” or through independent use).
This free workshop will be held in German.
Requirements
- Basic knowledge of Excel or a comparable spreadsheet application
- Excel or a comparable application installed on your computer
- Access to an AI assistant (e.g. ChatGPT, Copilot, Blablador, or Claude). If you do not have access to one, HIDA can provide you with a temporary ChatGPT account for the workshop. Further information will be provided after registration.
- If possible, an example of a data analysis task from your own work for which the result is already known or can be verified. If you cannot bring your own example, a sample dataset will be provided.
- A quiet workspace with a functioning camera and microphone, as the workshop includes several interactive working sessions
Course Dates
Places are limited to 20 participants and will be allocated on a first-come, first-served basis.
The Trainer
Ilona Lipp
Ilona Lipp
Dr Ilona Lipp works in the administration of Leipzig University as an Open Science Officer and is also a freelance trainer for research institutions, where she delivers workshops on a range of topics. One of her main areas of focus is the responsible use of generative AI.
After studying psychology, she spent more than twelve years conducting interdisciplinary neuroscience research in Austria, the UK, and Germany, working extensively with complex data analyses. She subsequently worked as a data scientist training neural networks. In recent years, she has undertaken extensive professional development in teaching and learning, coaching, and mediation to support learning, collaboration, and personal development in an effective and flexible way.
Stay up to date
Register for our AI in administration updates to be informed as soon as new courses go online.
