Building your AI agent with Agent Development Kit (ADK) and Postgres

In this hands-on workshop, you'll build a conversational AI agent from scratch and transform how you interact with your users.
You'll use Agent Development Kit (ADK) to orchestrate agent behavior and manage conversational flows, like handling complex customer queries about product features or stock. Then, we'll dive into powering your agent with the Postgres Vector Database and its pgvector extension. You'll gain practical experience with setup, vector embedding management, and advanced querying. This includes Vector Similarity Search for intelligent product recommendations (e.g., similar sportswear, complementary items for RAG), Full-Text Search for keyword-based discovery (e.g., product descriptions, reviews), and Location-Aware Insights for geo-spatial data (e.g., nearest store branches, in-store item location).
We'll seamlessly connect these data capabilities to your ADK agent using the MCP ToolBox. Through extensive live coding, you'll build and deploy a functional, data-driven conversational AI.
Key Takeaways: Understand core AI agent architecture with Google ADK, master pgvector for advanced multi-modal search in Postgres, and gain hands-on experience integrating a robust data backend with an AI agent to build your own working solution.
Target Audience: Developers and data engineers familiar with SQL and basic Python. Prior AI/ML experience is beneficial but not required.