Building a multi-agent system in a notebook is one thing. Building one that works reliably in production is another. This live 6-hour workshop teaches you the architectural discipline — context engineering, MCP, RAG, and safeguards — to bridge that gap.
By Packt Publishing · Refunds up to 10 days before
Multi-agent systems break in production because of context confusion, memory failures, and poor orchestration. This workshop teaches you to architect systems that solve these problems from the start — using semantic blueprints, MCP, and the Glass-Box Context Engine approach.
Context engineering is the discipline of designing AI systems that provide the right information, tools, and context to LLMs at the right time — replacing brittle prompts with reliable, scalable production AI architectures.
Multi-agent systems are AI architectures where specialised agents collaborate to accomplish complex tasks. This workshop shows you how to orchestrate them reliably using the Model Context Protocol and semantic blueprints.
MCP is Anthropic's open standard for connecting AI models to tools, data sources, and other agents. This workshop teaches you to use MCP for building orchestrated multi-agent workflows that are transparent and controllable.
Context engineering and multi-agent systems have almost no quality hands-on resources. This 6-hour live workshop gives you a complete guided build with a bestselling AI author answering your questions throughout.
Six modules. Six hours. A production-ready context engine by the time you finish.
Design structured context that gives AI agents precise, goal-driven contextual awareness beyond simple prompting.
Orchestrate specialised agents using the Model Context Protocol for adaptable, context-rich reasoning workflows.
Engineer retrieval-augmented generation pipelines with citations, memory, and safeguards against hallucination.
Design AI memory systems that maintain context across long conversations and complex multi-step workflows.
Implement moderation, data poisoning protection, prompt injection prevention, and trust mechanisms for production AI.
Build a transparent, traceable Context Engine that gives you complete visibility and control over your AI system.
A working production system — not just architectural knowledge.
A fully working multi-agent system with context engineering
MCP-orchestrated agent workflows you can use in production
High-fidelity RAG pipeline with citations and memory
Semantic blueprints and agent architecture patterns
Production-ready safeguards against hallucination and injection
Certificate of completion from Packt Publishing
Denis Rothman has 30+ years building AI systems and wrote the definitive book on context engineering for multi-agent systems.
Denis Rothman is a bestselling AI author with over 30 years of experience in artificial intelligence, optimisation, and agent systems. He has written multiple cutting-edge AI books for Packt Publishing and is the author of the book “Context Engineering for Multi-Agent Systems.” In this workshop he guides you step by step through the practical architecture of production-ready multi-agent AI systems.
This is an intermediate to advanced workshop. You need the basics below.
Common questions about the workshop, what to expect, and how to prepare.
You will build a Glass-Box Context Engine — a transparent, production-ready multi-agent system with MCP-orchestrated agent workflows, semantic blueprints for context design, a high-fidelity RAG pipeline with citations, memory engineering across conversation turns, and safeguards against hallucination and prompt injection attacks.
The hardest parts of building production multi-agent systems are: managing context across multiple agents without confusion, engineering reliable memory systems, building RAG pipelines that do not hallucinate, preventing prompt injection, and maintaining transparency into agent reasoning. This workshop addresses every one of these with concrete architectural solutions.
Yes. Denis Rothman structures this workshop to take you from first principles to a complete working production multi-agent system in 6 hours. The focus is on building a real system — not on comprehensive theory — so every hour of the workshop involves actual implementation guided by the instructor.
The workshop uses the Model Context Protocol (MCP) for agent orchestration, Python for implementation, and standard open source libraries for the RAG pipeline. The instructor covers why these choices are made and how the patterns translate to other frameworks like LangGraph, AutoGen, and CrewAI.
Yes. The Glass-Box Context Engine architecture you build is designed as a reusable template. The instructor covers how to adapt the semantic blueprints, MCP orchestration patterns, and RAG pipeline for different use cases and domains after the workshop.
Yes. The context engineering and MCP orchestration principles taught in this workshop are framework-agnostic. If you are already using LangChain, you will learn architectural patterns that improve your existing agent systems regardless of the framework. The instructor discusses LangChain integration during the session.
6 hours. Live bestselling AI author. Working production multi-agent system by the end. Seats are limited.
Register Now →Saturday April 25 · 9am to 3pm EDT · Online · Packt Publishing