In this tutorial, learn how to build, test, refine, and deploy a Python-based AI agent using the code authoring experience in Oracle AI Data Platform. See how an existing LangGraph project can be brought into the code editor, configured with Oracle AI Data Platform Utilities, and prepared to work with Oracle Cloud Infrastructure Generative AI models, guardrails, built-in tools, and project dependencies. The tutorial shows how to configure a retrieval-augmented generation (RAG) tool that retrieves relevant information from a knowledgebase, configure a language model, identify the project’s entry file, and use a requirements.txt file to manage third-party Python libraries. It also covers running the Python file to validate code, deploying the current agent version to Playground for conversational testing, reviewing execution traces to understand request processing, refining the code, and redeploying changes for validation. Finally, see how to select the target AI compute and deploy the agent for production use. Together, these capabilities provide one environment for developing, testing, and deploying existing Python-based AI agent projects.
For more information, see the documentation.
