Practical essays on product engineering, technical leadership, and the tradeoffs behind maintainable systems for focused teams.
Format
Engineering dispatches
Topics
Product / Systems / Teams
Archive
Latest first
Dispatch archive
Practical notes / Current index
01.
August 21, 2026
A software handover is complete when the receiving team can build, deploy, diagnose, and recover the system. This checklist turns access, documentation, and operational knowledge into verifiable acceptance checks.
02.
August 19, 2026
Software project rescue begins by restoring operational visibility and narrowing the work before new commitments are made. This guide separates stabilisation, essential delivery, and structural repair.
03.
August 15, 2026
Choose your first business automation by comparing frequency, value, stability, exceptions, system access, failure recovery, and ownership. The included scorecard keeps tool selection behind the process decision.
04.
August 14, 2026
Custom internal tools become useful when they complete one operational workflow with clear permissions, integrations, exceptions, and support ownership. This guide covers the shortest path from manual work to maintainable software.
05.
August 3, 2026
A useful codebase audit connects reproducible technical findings to the next business decision. This framework covers behaviour, security, architecture, operations, ownership, and prioritisation.
06.
July 28, 2026
A software project discovery phase should reduce the assumptions capable of changing the build decision. This guide shows how to investigate product and technical risk before committing to delivery.
07.
July 20, 2026
API-first AI agents become useful when a small interface constrains the data and actions available to the model. Server-side permissions, validation, and business rules provide enforceable guardrails around probabilistic behaviour.
08.
July 19, 2026
A startup’s technology choice affects how quickly future engineers can understand, change, and operate the product. This framework uses seven years of working with Go to explain why constrained, predictable technology often produces more maintainable software.
09.
July 18, 2026
A useful MVP scope defines what the product needs to prove, the smallest complete user workflow, and the technical requirements needed for a trustworthy test. This one-page framework helps founders set that boundary before development begins.
10.
July 17, 2026
A legacy rewrite should begin with a measurable target and a reliable way to compare old and new behaviour. This framework helps teams choose between retaining, refactoring, replacing, and rewriting before committing to a migration.
11.
July 16, 2026
A fractional tech lead helps founders turn an ambiguous product or engineering goal into a practical build path, then stays close enough to the code to help ship it. This guide compares the role with a fractional CTO, agency, senior freelancer, and full-time hire.
12.
July 15, 2026
AI coding agents are most useful when they absorb predictable implementation work. Visual HTML artifacts keep the work verifiable while engineers spend more time on product ambiguity, architecture, and decisions that are expensive to reverse.
13.
July 14, 2026
AI can make an internal tool quick to build, but the subscription comparison changes once colleagues depend on it. A useful build-versus-buy calculation includes expansion, maintenance, support, and the value of the work those hours replace.
14.
July 13, 2026
Coding agents make technical exploration and repeatable implementation cheaper. This practical agentic engineering workflow shows how to use that leverage while keeping architecture, domain understanding, and production responsibility inside the team.
Working principle / 01
Engineering decisions are product decisions when the team is small.
MBV Labs
Decide / Build / Learn
Editorial scope / Three fields
01.
Turning ambiguous goals into shipped systems.
02.
Clear choices for founders and lean teams.
03.
Architecture, operations, and code that can change.