Scoring Models and Prompts: Optimizing GenAI Decision Making

This 33-minute conference talk explores strategies for selecting the right LLM for your specific use case amid the rapidly evolving AI landscape. Discover Trellix's research framework for continuous model scoring that enables efficient use of different AI models by identifying their unique strengths and limitations for cybersecurity applications. Follow along as speaker Martin Holste, CTO of Cloud and GenAI at Trellix, shares real-world examples of using AI for critical decision-making and learn key approaches for developing your own sustainable model and prompt management strategy. Gain valuable insights into evaluating model performance, cost-effectiveness, and optimization techniques that can be applied across various industries and use cases.