Last month, I had the fantastic opportunity to run a hands-on Model Context Protocol (MCP) workshop. It was a chance to mentor and coach colleagues, from engineers and testers to product owners, as we explored what MCP is, how it works, and why it is generating so much momentum across AI and developer communities.
My goal was not just to share knowledge. It was to guide the learning journey and build a shared, foundational understanding of this powerful emerging standard.
Workshop Highlights
We covered everything from core protocol basics and prompt crafting to real world patterns such as:
MCP standardises how AI applications communicate with external systems using JSON‑RPC 2.0 over stateful, bidirectional connections. Unlike standard JSON‑RPC, which treats every request independently, MCP augments the protocol by embedding session tokens and context IDs into each message.
This enhancement provides state, enabling advanced, multi‑step interactions and dynamic module loading at runtime. Such a stateful design is essential for modern, agile AI systems.
Imagine an AI‑powered payments system that dynamically integrates a new fraud detection module during operation. MCP allows the system to load this module on the fly without a full redeployment, provided the server supports dynamic module loading.
In this post, we'll explore the MCP protocol, its core components, and how it compares to traditional REST and GraphQL APIs.
AI applications are shifting from monolithic large language models to modular, multi-agent systems—a transformation that enhances performance, flexibility, and maintainability.
In my talk about AI Agents last September, I said AI agents would become increasingly popular. Today, we see this shift happening across industries. By breaking down complex tasks into specialised components, engineers can design smarter, more scalable AI workflows.
Analogy: Think of multi-agent systems like a well-coordinated orchestra. Each musician (agent) has a specific role, and together, they create a harmonious performance. In software, this means dividing complex problems into manageable, specialised parts that work in concert.
In this guide, we'll explore four key multi-agent patterns, using travel booking as an example. You'll learn how to choose the right pattern for your application, implementation strategies, and error-handling techniques to build robust multi-agent AI systems.
Agentic systems, where multiple AI agents collaborate through decision-making and handoffs, shine in specific scenarios but add operational complexity.
In this post, we'll explore the scenarios where agentic systems are most effective and the challenges you may face when using them.
It's like a new diet pill has hit the market, promising instant weight loss with no effort, and now everyone's scrambling to get their hands on it because it's all over their TikTok.
Some folks are talking about DeepSeek as if it's the second coming of AI.
The clamour might be because it's the first serious open non-US model with reasoning capabilities from China.
Most people are confused about which version of DeepSeek they're using.
Most providers offer a distilled, watered-down version that'll run anywhere, so you're not getting the actual full fat version people are talking as to really see its magic, you'd need a small fortune in computing power.
It's like being promised a rare vintage white wine, only to find the bottle filled with slightly grape-scented water. Same brand, same label, but all the depth and character stripped away, leaving you with a hollow imitation.
The hype suggests it can keep pace with anything OpenAI does, which might sound thrilling if you're already tired of whatever ChatGPT or its siblings spit out.
But here's where the alarm bells should start ringing.
Building AI products with stakeholders requires a fundamentally different approach than traditional software development.
From my experience working with AI, success depends not only on technical implementation but also on bridging the gap between AI capabilities and stakeholder expectations.
In this post, I’ll share lessons from guiding stakeholders through AI’s possibilities and limitations.
The goal of creating something "predictable," reliable, and consistent is a shared principle across all the teams I've worked with throughout my career.
Knowing that the same code would always return the same output when given the same inputs was the foundation of everything we built.
We aimed for no surprises, no matter how complex a workflow might be.
Whether implicitly or explicitly using finite state machines, this determinism enabled us to build testable, monitorable, maintainable, and, most importantly, predictable workflows.
We read and shared ideas at conferences, promoting patterns and principles like SOLID and DRY to create functional, composable, and extensible software.
Having lived through the era of a "new JavaScript framework every week," we now find ourselves in the gold rush of the AI agent framework space.
New frameworks appear daily, each claiming to be the 'ultimate' solution for building AI agents, often backed by YouTubers enthusiastically promoting demoware and usually their own library, framework, or SaaS offering. Unfortunately, this enthusiasm can lead companies to uncritically adopt these tools without considering the long-term implications.
