Serg Masís

Agentic AI with GraphRAG and MCP tools use

Learn to create a self-reflective agent with open source tools and your choice of foundation model (local, OpenAI, Claude, Gemini) based on hardware and preferences. The example we will use is of creating a chatbot to ask questions about your grandmother’s handwritten recipes. An ETL (extract, transform, and load) pipeline will interpret the recipes using a multimodal LLM and store them in a database, both graph and vector databases. And, then we will create a CrewAI chatbot agent that will query this database using a GraphRAG (Graph-enabled Retrieval Augmented Generation) when asked questions about the recipes and will use MCP (Model Context Protocol) tools to perform additional tasks with the recipes. During this experience, we will learn about functional Agentic AI patterns, frameworks, evaluation/testing methods, and other considerations. We will also learn about ETL, embeddings, APIs, and different foundation models.

Data Scientist in agriculture with a background in entrepreneurship and web/app development, and the author of the book "𝐼𝑛𝑡𝑒𝑟𝑝𝑟𝑒𝑡𝑎𝑏𝑙𝑒 𝑀𝑎𝑐ℎ𝑖𝑛𝑒 𝐿𝑒𝑎𝑟𝑛𝑖𝑛𝑔 𝑤𝑖𝑡ℎ 𝑃𝑦𝑡ℎ𝑜𝑛". Passionate about ML interpretability, responsible AI, behavioral economics, and causal inference.