Trustworthy GenAI: a practical playbook for industrial applications

Generative AI is moving quickly from experimentation into real industrial and digital applications, but making a GenAI system "work" is not the same as making it trustworthy. In this one-hour webinar, Sirris and Flanders Make will present the main findings of the TrustGenAI project, developed within the Flanders AI Research programme.

You will discover the Trustworthiness Playbook for GenAI, learn how to assess and reduce risks, and see how the approach is applied through two industrial use cases: grid power forecasting and multimodal operator support. By the end of the webinar, you will have a practical framework to strengthen the trustworthiness of your own GenAI applications.

For whom?

This webinar is intended for professionals who design, integrate, deploy or operate GenAI applications in an industrial context.

• FAIR (Flanders AI Research) researchers
• Engineers and technical teams
• Product teams and project managers
• GenAI practitioners
• Industrial decision-makers
• Non-technical profiles who want to better understand trustworthy GenAI

Key Takeaways

You will get a practical approach to assess and strengthen the trustworthiness of GenAI applications. The focus is on technological innovation that is safe, useful and relevant for industry.

• Identify the nine main trustworthiness aspects for GenAI applications.
• Learn which aspects are must-haves and which are nice-to-haves for a specific use case.
• Understand why trustworthiness depends on the context, not only on the model.
• Discover concrete design choices to reduce risk.
• See the approach applied to two industrial use cases: grid power forecasting and multimodal operator support.
• Use the playbook as a starting point to design, evaluate or improve your own GenAI applications.

Agenda

Language: English