A provisional page for my workshop at PyConES 2026 in Barcelona.
LangGraph makes it possible to build controllable, maintainable agents. Its full potential emerges when we combine it with knowledge systems and information retrieval.
In this hands-on workshop we will build an agent step by step using Python and LangGraph, starting with a basic graph and progressing to a system that can retrieve external knowledge through RAG and evolve with use.
Workshop details
- Speaker: Miguel Ángel Fernández Gutiérrez.
- Conference: PyConES 2026, Barcelona.
- Date: Friday, 6 November 2026.
- Time: 18:00–19:30.
- Duration: 90 minutes.
- Format: hands-on workshop.
- Track: machine learning and artificial intelligence.
- Room: Talleres aula 05.
- Language: Spanish.
What we will build
During the session we will learn to model state, define nodes and control execution flow. We will then add a hybrid retrieval system so that the agent can access relevant information in real time.
Finally, we will explore how to introduce mechanisms for “learning”: storing new knowledge and reusing it in future interactions.
What you will take away
By the end of the workshop, you will:
- Understand how to build LangGraph agents from scratch.
- Know how to integrate RAG in a structured way, beyond simple examples.
- See how to design agents that evolve through persistent knowledge.
- Have a solid foundation for building real systems in production.
Who it is for
Developers with basic Python knowledge who are interested in applied AI systems.
