Agentic Engineering with Jev
Jev is emerging as a new piece of the agentic AI stack. Learn where it fits alongside LLMs and build a complete AI agent using Jev, Python, Tools and MCP.
Build Reliable AI Agents with Jev, Python, LLMs, Tools & MCP
Learn how Jev fits into the modern AI agent stack and build a complete agentic system from scratch.
As AI agents evolve, developers need to understand not only how to use LLMs, but how to combine them with lighter models, tools, MCP, coding agents, and structured agent workflows.
In this hands-on workshop, Jaime Buelta will show you how to use Jev alongside LLMs to build faster, more structured, and more reliable AI agent workflows.
What You'll Learn
Understand what Jev is and where it fits in an AI agent architecture
Learn how Jev and LLMs can work together inside agentic workflows
Build AI agents using Python, LLM APIs, tools, memory, and agent loops
Work with local and hosted models
Build and connect your own MCP server
Extend Claude Code or Codex using MCP, skills, and external tools
Move beyond vibe coding with spec-driven and reviewable development
Control how and when agents use tools
Test, debug, inspect, and audit agent behaviour
Improve agent reliability and handle common failures
Build a complete end-to-end AI agent as your capstone
Workshop Modules
Module 1 – Understanding AI Agents: LLMs, Jev and MCP
What an AI agent is and how it works
LLMs versus System 1 / Jev-style models
Local vs. hosted models
Context and agent loops
Understanding MCP protocol and other tools
Real world agents. General vs specific
Module 2 – Agentic Coding that works
Working with an agentic coding tool such as Claude Code or Codex
MCP servers, skills, and extending coding agents
From "vibe coding" to solid Agentic engineering
Module 3 – AI Agent blueprints with MCP and Jev
Understanding the components that make up an agent
Calling an LLM through an API, both remote and local
Giving the agent tools and controlling how it uses them
Using Jev decisions and categorisations
Creating and connecting an MCP serverMake the agent to work on real life
Module 4 – Capstone: Build an Agentic Application
Bring LLM, MCP, Jev and other tools together into a working project
Handle failures and reliability
Understand agent behaviour and iterate
Experiment and extend the application!
Why This Workshop Now?
Most AI agent workflows still rely heavily on large language models for every task.
But not every step requires an LLM.
You'll explore how Jev can act as a fast System 1 model alongside LLMs, helping you think differently about how modern agent workflows can be designed.
The goal is simple:
Know when to use Jev.
Know when to use an LLM.
Know how to connect them into a reliable agent system.
Who Should Attend?
Ideal for:
Software Developers
Python Developers
AI / ML Engineers
AI Agent Builders
Claude Code / Codex users
Developers exploring MCP
Engineers looking to move beyond basic LLM experimentation
By the End of the Workshop
You'll have practical experience building an agent that combines:
Jev + Python + LLMs + Tools + MCP + Agent Logic and you'll understand how to test, inspect, debug, and improve it, rather than simply hoping it works.
Stay ahead of the Jev wave and learn how to engineer it into real AI agent workflows.
Meet Your Instructor:
Jaime Buelta is a software architect, developer, author, and long-time Python practitioner.
His work spans Python development, software architecture, automation, microservices, and building scalable software systems. He is the author of Packt titles including Python Automation Cookbook and Python Architecture Patterns. His latest work around Python automation also covers AI models, MCP integrations, and the creation of AI agents.
Throughout this workshop, Jaime will bring an engineering-first perspective to AI agents focusing not just on what agents can do, but how their underlying systems should be designed, inspected, and improved.