Blog

What is my team missing? A team only hears what it has chosen to listen to. Five places to listen, where AI helps, and a simple test of balance for leaders.

Generated reports cost nothing to write and a reader's attention to read. One practice to try: name the reader, ask for something, and treat action as proof.

AI makes polished tickets, replies and pull requests cheap, so they no longer prove understanding. A test for which delivery practices still create it.

AI adoption keeps climbing but measurable returns mostly do not. The gap is discipline, not models: living references, scoped roles and firm review gates.

Cloud AI pricing is subsidised and will rise. This post argues for choosing local, on-premises or cloud AI by workload, with open-weight models now viable.

Career ladders reward production seniority, but AI agents now produce. What a new ladder must measure, and why entry-level roles need redesign, not removal.

The AI-era engineering team needs an operating model, not a new org chart. This post shows where AI executes, where humans judge, and what each role now does.

AI has cut the cost of writing code, not maintaining it. This post argues for exploring more before committing, as judging what to keep is the bottleneck.

Boards hear AI will write 90% of code and cut engineering headcount. Why code is not engineering, and what AI amplifies and what it cannot touch.

AI agents lose track of multi-step work because they keep process state in chat context. The storage layers they lack, and the workarounds teams use today.

Most organisations run AI without data governance, and the EU AI Act adds legal risk. Covers shadow AI, agent data access and a phased path to compliance.

Firms spend billions on AI while discounting senior engineers. This post argues the 'Junior + AI = Senior' bet is bubble behaviour and will correct.

AI coding agents default to markdown and CSV where a database would do. This post shows why, and how to prompt agents towards the right data store.

Most teams know what to measure about testing but cannot see enough to do it. Five visibility gaps, from execution tracking to cross-team data.

AI adds new failure modes to incidents and faster ways to respond. This post covers AI-specific runbooks, recovery basics and a four-phase adoption plan.

Most teams do not need another AI tool. Start with markdown, Git and CLI tools, then use five questions before buying observability, orchestration or MCP.

Treating Claude Code like a new engineer, with discovery, living docs, task breakdown and governance, delivered an event-driven system without major rework.

The 2025 DORA report finds AI amplifies the organisation you already have. The seven capabilities that matter, and why speed gains outrun stability.

AI agents lose context across sessions. External memory files, handoff notes, checkpoints and recovery steps keep long multi-phase projects on track.

Coverage does not show whether tests work. Measure test effectiveness through execution patterns, flakiness, failure analysis and local versus CI differences.

Six testing antipatterns that AI-assisted development exposes, from skipped test-first work to coverage theatre and flaky tests, with questions to spot them.

Why testing discipline matters more when AI writes the code: how test-first work guides AI, how the velocity trap catches teams, and how to keep verifying.

How non-developers can use Claude Code, a command-line AI agent, for meeting notes, reports and research, using slash commands, hooks and a prompt library.

Vague AI requests give poor results. How to break user stories into bounded, prompt-ready work units with clear inputs, outputs and standards.

Copying docs into CLAUDE.md files makes every AI query costly. The Compass Pattern points agents to existing docs, cutting input tokens by 92% in one project.

Teams code faster with AI but ship at the same speed. This post argues org structure is the constraint, and sets out new roles, team shapes and a roadmap.

Seniority is judgement, systems thinking and influence, not years or titles. How senior engineers lead without authority, and why AI cannot replace that.

AI-assisted juniors look productive but get promoted beyond their ability. Why career timelines still matter, and how to mentor when AI writes half the code.

A junior engineer plus AI looks cheaper than a senior, but hidden costs and skills gaps reverse that. This post argues for AI spend that builds capability.

What the miniature painting community's row over AI at Golden Demon teaches engineering leaders about skill growth, authenticity and drawing lines on AI.

Firms are freezing engineering hiring while waiting for AI to replace engineers. This post argues that is a strategic error, and what to do instead.

Treat decisions like code changes and write them as pull requests. It stops repeated debates, keeps context and gives AI assistants the history they lack.

How engineering leaders can capture, triage and act on a constant stream of ideas, using RICE, the Eisenhower Matrix and AI tools, without losing the best.

AI amplifies the communication habits a team already has. Why clear written communication comes first, with frameworks for requirements, feedback and AI use.

Why reliability must come before features, and how to make operational excellence a shared responsibility across the organisation, not one team's job.

Split AI context into fixed, session and variable layers and keep each in a markdown template, so you stop re-explaining yourself and get steadier results.

Coding interviews that ban AI test the wrong skills. Assess how candidates work with AI, critique its output and judge when not to use it.

Why the usual objections to hybrid mobile apps on performance, user experience and native features are out of date, and why team expertise decides results.

Why per-token, per-plan AI pricing discourages trying more than one model, hides the cost of experimenting and muddies ROI, plus ways to budget by outcome.

What changes when AI moves from assisting to acting: agent problems with state, tooling and accountability, and what AI-native architecture asks of teams.

AI will not write 90% of your code, but it will change what you build. Part one of two on bolting AI onto existing systems and what breaks when you do.

Story points and velocity mislead. Throughput, cycle time, lead time, flow efficiency and quality show how work flows, with eight steps to move your team.

PI Planning, story points and velocity promise predictability but breed false certainty and gaming. This post explains why, and how planning for flow helps.

Why story point estimation gives false precision, eats planning time and rewards effort over value, and why velocity says little about what customers get.

Why Wyrd Technology rebranded: the Norse and Saxon idea of wyrd as a web of connected outcomes, and what the new logo, website and blog are for.