In this session you are going to learn how to build an AI Agent from scratch using the Vertex AI SDK for Python and osquery. This agent will enable you to run diagnostics in a machine using natural language, enabling you to ask questions like "what are the top memory consuming processes?" or "can you find any signs of malware on my system?"

This is made possible because LLMs are particularly good at writing SQL queries and osquery enables us to query operating system information through SQL. By combining both technologies in an agent interface we can have human-like interactions with our machine, potentially speeding up the diagnostic process. Is this the beginning of "AgentOps"*?

Topics covered in the session:

• What is an agent?
• What is osquery?
• Agent-model interactions
• Fine-tuning the agent
• Session control
• Deploying the agent to the cloud

(*Not really - unfortunately other people already claimed this term! :)