LLM and RAG evaluation: Detecting vulnerabilities using KNIME workflows with Giskard

Discover how to build secure AI applications by identifying vulnerabilities in your LLM and RAG systems using KNIME's workflows and Giskard's evaluation capabilities. In this session, you'll learn practical strategies to enhance the reliability and security of your LLM-based applications, from development to deployment.

Key takeaways from this session:

• Explore a real-world LLM and RAG use case using KNIME's visual workflow environment
• Learn how to implement Giskard's LLM Scanner and RAG Evaluator within KNIME
• Understand common LLM vulnerabilities and their impact on AI systems
• Discover strategies for improving AI workflow security and performance

Speakers:

Jean-Marie John-Mathews - Co-CEO and co-founder at Giskard

Jean-Marie holds degrees in Data Science (ENSAE), and in Philosophy and Quantitative Economics, with a PhD in AI Ethics from Université Paris-Saclay. His expertise in data science, AI Ethics, and user research is crucial in ensuring Giskard's product meets user needs.

Roberto Cadili - Data Scientist, Evangelism team at KNIME

Roberto has an MSc. in Social and Economic Data Science from the University of Konstanz, specializing in machine learning, NLP, and Computer Vision. As editor for Low Code for Data Science, he helps the KNIME community develop and share successful data science practices.

Join us and KNIME to gain these skills for ensuring you can confidently deploy more secure and trustworthy LLM and RAG systems 🚀