9 min read
Why LLMs ignore the middle of your context
Long context windows do not guarantee reliable recall. Design retrieval, reranking and prompt assembly so the model can use the evidence it receives.
Blog posts tagged
ARCHITECTUREPage 1
5 min read
Configure AI Models Instead of Hard-Coding Them
Keep model names in configuration, select models by task, and compare routes using the cost of completed work.
8 min read
The Laravel action pattern deserves to escape Laravel
Why modelling each business operation as a reusable action keeps controllers, jobs, commands, listeners and other delivery mechanisms thin.
7 min read
Why Single Agents Often Beat Multi-Agent Systems
Why many apparent multi-agent gains are really test-time compute gains, and when extra agents are still worth the complexity.
7 min read
Agentic UI: Comparing AG-UI, MCP-UI, and A2A Protocols
A practical comparison of three emerging agent protocols, what layer each one solves, and when to use them together.
8 min read
How LLMs Come Into Being: A Beginner-Friendly Guide to Data, Tokens, Training, and Inference
A beginner-friendly but detailed guide to how large language models are built, trained, aligned, and run in production.
6 min read
Parallel State Machines: Modelling Independent Behaviour Without a Giant FSM
When one workflow is not enough, parallel state machines keep independent concerns separate, testable, and easier to reason about.
7 min read
State Machines in Practice: Advantages, Disadvantages, and a TypeScript Example
A practical, engineer-friendly look at state machines: why they’re great, where they hurt, and a clean TypeScript finite-state-machine example.