Jan 17
Loading video player...
Use case: Multi-agent Agentic AI system - External data source tracker Chapters 00:00 Introduction 00:37 About the use case (One main agent, 3 sub agent - ClinVar, dbNSFP and Gnomad specialist) 04:39 Explore the 3 sub agent tasks, the database page to fetch release notes 07:38 Tools needed for parsing database 08:12 Hands on, start with environment setup 09:42 Define tools 13:17 Define 3 sub agents (ClinVar, dbNSFP and Gnomad) 15:44 Convert sub agents as tools 16:42 Define main agent 17:16 Interact with agent in a chat interface 18:19 Explore the traces 27:45 Alternate Design patterns from openAI documents page Code: https://github.com/prpanigrahi/ai_agents_bioinformatics_course/blob/main/basics/4_multi-agent.ipynb In this video, we build a real multi-agent AI system for bioinformatics that tracks release updates from multiple biological databases and summarizes them automatically. This is the seventh video in the series “Building AI Agentic Applications for Life Science”, and it introduces true multi-agent reasoning and coordination. 🧬 What this system does: • Fetches latest release updates from ClinVar, gnomAD, and dbNSFP • Uses specialized agents for each database • Uses a master agent to decide which agent to call • Iteratively reasons and acts using ReAct-style loops • Summarizes updates into a clean markdown format • Exposes everything through a chatbot interface 🧠 Key concepts covered: • Multi-agent architectures • Master–worker (master–slave) pattern • Agent specialization • ReAct-based reasoning loops • Agent-to-agent orchestration • Chat-based interaction with multi-agent systems ⏱️ Video length: ~32 minutes 🚀 Who is this for? • Bioinformaticians • AI engineers building agent systems • Researchers interested in autonomous AI workflows 👉 This video shows how complex scientific workflows can be automated using multiple cooperating AI agents. 👍 Like | 💬 Comment | 🔔 Subscribe for more advanced agentic AI use cases