Overview

In this full-day workshop, participants build on the "Oracle AI Data Platform Essentials" curriculum and dive into Oracle AI Data Platform’s new agentic capabilities. After a short review of foundational features, they will learn key concepts of AI agent creation. They will then move into hands-on labs to build an AI agent through code and through the visual builder interface, and surface it through an interactive chat UI.

Participants will learn how to:

• Build an end-to-end Agent Flow that chains MCP, Retrieval-Augmented Generation (RAG), SQL , HTTP, and Prompt, tools.
• Integrate OCI Generative AI foundation models to drive AI Agent interactions.
• Implement guardrails for content moderation, PII, and prompt injection.
• Create Knowledge Bases for integration with RAG Tools and semantic retrieval.
• Integrate SQL Tools for running parameterized queries against registered external catalogs.
• Use the MCP Server tool to integrate AI Agents with external MCP servers.
• Use the HTTP tool to call endpoints, allowing agents to take actions and receive information.
• Create Prompt Tool templates with parameters to create specific output.
• Build an end-to-end Agent Flow with all these features using code, or with the visual builder interface.

Attendees will also learn how to operate and govern:

• Knowledge Base and Agent Flow permissions.
• AI Compute clusters for agent development and testing, including creating, sizing, and attaching clusters.
• RBAC for AI Compute clusters to control who can attach, deploy, or modify resources.
• Agent Flow deployment to AI Compute, including retrieving usage metrics for monitoring.
• Agent Flow deployment URIs for serving the deployed model.
• Agent Flow monitoring and troubleshooting.

Hands-on labs guide participants through each step, with scheduled Q&A sessions throughout.

Featured speakers

Miles NovotnySenior Product Manager, Analytics

Joyce OfiliPrincipal Product Manager, Analytics