Building Agentic Applications: A Deep Dive into Zooka, an AI Cardiologist Assistant

Building a basic chatbot is easy. But engineering a true agent that reasons, remembers, and acts often feels like a massive undertaking. We will start with the fundamentals: exploring the architecture and the building blocks of agentic applications and comparing the different abstraction frameworks available to build them. Once we have the theory down, we’ll put it into practice with Zooka, a fully functional AI Cardiologist Assistant. You’ll see how efficient modern development can be as we walk through Zooka’s core logic, written in under 300 lines of code. Finally, we’ll look under the hood at how this architecture handles complex challenges like session management and user-scoped long-term memory. We’ll cover data integration using the Model Context Protocol (MCP) for vector and knowledge graph searches, and show exactly how to wire it all up to a live frontend application