ALC Workshop 1 - Building a Secure AI Agent

This workshop teaches attendees about the basic technology of AI models, and has practical exercises to get hands-on with running the tokenization and embedding functions used in AI models. It explains the development of large language model (LLM) Generative AI as a newly emerged disruptive technology, and outlines with practical demonstrations some of the threats to AI models and the problems that can occur when AI models run. It then shows through presentations and demonstrations how guardrails can be used to mitigate these threats. The workshop then progresses to the design of AI Agents which integrate both LLM and tools to create an AI agent that can take actions in the real world and has protection by use of a built in guardrail.

Attendees can join the instructor in running a set of 6 labs to build the tools and the AI Agent.