Uploaded August 2026 | Updated September 2026, 2 weeks ago
Build governed multi-agent systems in Oracle AI Data Platform to coordinate specialized AI agents, automate business processes, and validate agent behavior. See how Visual Builder, guardrails, conversation memory, traces, and testing support secure AI development, at the additional resources linked below.
In this tutorial, see how to build, test, and deploy a multi-agent system using Visual Builder in Oracle AI Data Platform. The tutorial shows how a supervisor agent can coordinate specialized executor agents to automate an expense compliance workflow, including evaluating expense reports against company policies and answering spending and budget questions with an Oracle Analytics Cloud connection. See how to connect agents and other components on the visual canvas, configure guardrails to block potentially unsafe content and prompt injection attempts, and manage how conversation history is shared across agents.
The tutorial also covers private state isolation, which maintains separate conversation histories between executor agents and the supervisor agent, as well as limiting conversation history to the last 20 messages. Testing in the Playground helps validate the complete multi-agent workflow or individual agents independently. Traces provide visibility into how the supervisor agent delegates tasks to executor agents and assembles responses, while guardrail testing verifies protections for incoming messages and generated responses.
The tutorial concludes by showing how deployment makes the agent available for production use on the selected AI compute and provides endpoint URLs for communicating with the deployed agent. Together, these capabilities support governed, collaborative multi-agent applications through a visual, low-code development experience.
Find training paths for Oracle Analytics and AI from beginner to advanced and expert: https://social.ora.cl/6006B1KD0j
Check out detailed information on AI Agents: https://social.ora.cl/6000B1KKdp
Like what you learned about Oracle Fusion AI Data Platform? Subscribe now https://social.ora.cl/60025k87j
Join us at Oracle AI World: https://social.ora.cl/60035eVvJ
Contact AIDP sales: https://social.ora.cl/6003FLUMT
Learn more about Oracle AIDP: https://social.ora.cl/60047h76Q
Join the Oracle AIDP Community: https://social.ora.cl/60087h7Bj
Follow us here:
LinkedIn: https://social.ora.cl/60035eXuF
X: https://social.ora.cl/6008abVgG
Instagram: https://social.ora.cl/60095eXDN
Facebook: https://social.ora.cl/60065enQp
Build governed multi-agent systems in Oracle AI Data Platform to coordinate specialized AI agents, automate business processes, and validate agent behavior. See how Visual Builder, guardrails, conversation memory, traces, and testing support secure AI development, at the additional resources linked below.
In this tutorial, see how to build, test, and deploy a multi-agent system using Visual Builder in Oracle AI Data Platform. The tutorial shows how a supervisor agent can coordinate specialized executor agents to automate an expense compliance workflow, including evaluating expense reports against company policies and answering spending and budget questions with an Oracle Analytics Cloud connection. See how to connect agents and other components on the visual canvas, configure guardrails to block potentially unsafe content and prompt injection attempts, and manage how conversation history is shared across agents.
The tutorial also covers private state isolation, which maintains separate conversation histories between executor agents and the supervisor agent, as well as limiting conversation history to the last 20 messages. Testing in the Playground helps validate the complete multi-agent workflow or individual agents independently. Traces provide visibility into how the supervisor agent delegates tasks to executor agents and assembles responses, while guardrail testing verifies protections for incoming messages and generated responses.
The tutorial concludes by showing how deployment makes the agent available for production use on the selected AI compute and provides endpoint URLs for communicating with the deployed agent. Together, these capabilities support governed, collaborative multi-agent applications through a visual, low-code development experience.
Find training paths for Oracle Analytics and AI from beginner to advanced and expert: https://social.ora.cl/6006B1KD0j
Check out detailed information on AI Agents: https://social.ora.cl/6000B1KKdp
Like what you learned about Oracle Fusion AI Data Platform? Subscribe now https://social.ora.cl/60025k87j
Join us at Oracle AI World: https://social.ora.cl/60035eVvJ
Contact AIDP sales: https://social.ora.cl/6003FLUMT
Learn more about Oracle AIDP: https://social.ora.cl/60047h76Q
Join the Oracle AIDP Community: https://social.ora.cl/60087h7Bj
Follow us here:
LinkedIn: https://social.ora.cl/60035eXuF
X: https://social.ora.cl/6008abVgG
Instagram: https://social.ora.cl/60095eXDN
Facebook: https://social.ora.cl/60065enQp
