Skip to main content

Writing

Notes on AI, agents, web technologies, and building with LLMs

Technical articles on building with LLMs: agents and agent harnesses, MCP, evaluation, and the web technologies underneath. Worked examples with code, written for working developers.

The newsletter

Ctrl+Alt+Delegate

The three-finger salute for work you should not be doing yourself.

  1. MCP Servers - The Bridge Between LLMs and Real-World Tools

    Model Context Protocol (MCP) servers expose tools, resources, and prompts to LLMs in a unified, structured way. This post explores how they work, how to build one, and why they are a critical part of the future AI stack.

  2. Consuming Streamed LLM Responses on the Frontend: A Deep Dive into SSE and Fetch

    Learn how to build a responsive, real-time user experience by consuming streamed Large Language Model responses on your frontend. This article provides a comprehensive guide to using both Server-Sent Events (SSE) and the Fetch API with Readable Streams, complete with code examples and a detailed comparison.

  3. Filling in the Blanks: Teaching AI to Inpaint

    A hands-on guide for exploring how to train a simple AI model using TensorFlow.js to inpaint missing parts of images - without needing large datasets or prior machine learning experience.

    InteractiveLLMAI
  4. Agentic AI: Multi-Agent Systems and Task Handoff

    The final article in the Agentic AI series explores multi-agent systems: how specialised agents collaborate through structured handoffs to complete complex user goals.

    InteractiveAgentic AI
  5. Understanding the Orchestrator-Worker Pattern

    The orchestrator-worker pattern brings scalable structure to agentic AI workflows by cleanly separating high-level planning from specialised task execution. Through a practical trip planning example, this article demonstrates how LLMs can dynamically coordinate expert agents, grounded in schema-driven logic and real-world data.

    InteractiveAgentic AI
  6. Building with Reflection: A Practical Agentic AI Workflow

    This article explores how to implement a reflection loop-an agentic AI pattern where a model generates, critiques, and iteratively improves its output - using image captioning as a practical example.

    InteractiveAgentic AI
  7. Parallelisation as an Agentic Workflow

    Unlock faster, more diverse reasoning by running multiple LLM prompts in parallel and aggregating their responses into a single, cohesive output.

    InteractiveAgentic AI
  8. How Transformers and LLMs Actually Work - A Developer's Guide with Code

    A hands-on walkthrough for web developers to demystify large language models by actually building a mini Transformer from scratch.

    InteractiveLLMAI
  9. Routing: Building Step-by-Step AI Reasoning

    Explore how to intelligently route AI queries using schema-guided function calling and contextual categorisation.

    InteractiveAgentic AI
  10. Prompt Chaining: Building Step-by-Step AI Reasoning

    Learn how prompt chaining enables AI to tackle complex tasks through step-by-step reasoning, boosting both accuracy and interpretability.

    InteractiveAgentic AI