Getting a multi-agent AI system to work in a notebook is straightforward. Getting it to work reliably in production is the hard problem. This live 6-hour workshop teaches you to solve it — with context engineering, MCP, high-fidelity RAG, and production safeguards.
By Packt Publishing · Refunds up to 10 days before
Demo AI systems work because they run on clean data, short conversations, and happy paths. Production AI systems must handle noisy data, long contexts, adversarial inputs, and failure modes. Context engineering is what bridges this gap — and this workshop teaches you how.
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 for production — not just for academic papers and blog posts.
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.
A production-ready multi-agent AI system has: transparent and traceable agent reasoning (Glass-Box), reliable context management that prevents overflow and confusion, high-fidelity RAG that cites sources and prevents hallucination, safeguards against prompt injection and data poisoning, and graceful error handling. This workshop builds all of these from the ground up.
The most common production failures are: context window overflow causing agents to lose important information, context confusion when multiple agents share information poorly, hallucination in RAG responses, prompt injection attacks that hijack agent behaviour, and coordination failures between agents. Context engineering addresses all of these systematically.
This 6-hour workshop takes you from first principles to a complete production-ready multi-agent AI system. The longer format compared to typical 4-hour workshops is intentional — production AI architecture requires time to implement properly, not just understand conceptually.
The workshop follows a structured build of the Glass-Box Context Engine architecture. If you have an existing project, the principles and patterns you learn will directly apply. The instructor can discuss how to adapt the architecture to specific use cases during the Q and A portions of the session.
Documentation covers what APIs do. This workshop covers how to architect systems that use those APIs reliably — semantic blueprint design, context window management strategies, RAG validation approaches, memory engineering patterns, and safeguard implementation that you will not find in framework documentation.
Yes. The Glass-Box Context Engine architecture taught in this workshop is specifically designed for enterprise requirements — auditability, transparency, controllable behaviour, and production reliability. Many enterprise AI teams attend this workshop to establish architectural standards for their agentic AI development.
6 hours. Live bestselling AI author. Production-ready multi-agent AI by the end. Seats are limited.
Register Now →Saturday April 25 · 9am to 3pm EDT · Online · Packt Publishing