Engineering

Newsletter

From Andreas' Desk

A bi-weekly letter from our CEO Andreas Creten for SaaS leaders and CTOs: what we see in audits, AI engineering and the companies we work with.

You're in — check your inbox to confirm.

From our SaaS audits

What we find inside SaaS companies

Patterns from the technical due diligences we run for investors and boards.

See the audit findings

The testing pyramid scheme

Feature tests won because unit tests cost too much to write by hand. Agents changed that maths, but they copy whatever your suite already does. Rebuilding one app's suite took 128 HTTP tests down to 31, with 126 unit tests running in under a tenth of a second.

Press m for madewithlove

Moving one overloaded ~/.claude folder into three Claude Code accounts is mostly config: CLAUDE_CONFIG_DIR, the hidden ~/.claude.json, plugin paths and a logout. A small zsh function then asks which account to use, and resume finds the right one on its own.

Teenage Mutant Ninja Tests

Mutation testing was always too slow to run on a whole codebase, so nobody did. Agents change that: they touch a handful of files, don't mind a two-minute run, and escaped mutants tell them exactly which tests check nothing. A targeted Infection setup for PHP, plus the agent instructions.

How LLMs actually work (explained like I'm 5)

An LLM does one thing: it predicts the next token. Training nudges that guess towards the right answer billions of times, attention lets it look back at the tokens that matter, and chaining guesses together turns a prediction into an essay.

Engineering is no longer a distraction-free job, deal with it

Running agents in parallel turned engineers into notification processors. Defending focus is a losing fight. The job now is making interruptions affordable: visible signals, no friction, finished work you can walk away from, and more valuable work within reach.

Why DDD is more important than ever

AI made code cheap, so the bottleneck has moved from how to build to what to build. Strategic DDD matters more than ever, whilst tactical patterns like aggregates matter less as AI writes and reads more of the code.

Owning the software factory

My first software factory flooded the team with pull requests nobody asked for. The second one had the same problem, just with better visibility. It only started working when the team took ownership and the factory began to encode what we believe good engineering looks like.

Relive your childhood memories, now with no bugs

As a kid, your skills could only get you so far in a game. Sometimes the game itself was the problem. AI fixed a decades-old goalkeeper bug from a short video, and the real shift is what that means for anyone with an idea.

From template tag to hydrated component

A React island inside a Django page needs three things: a template tag that records rather than renders, one batched call to a render sidecar, and a browser that hydrates only the IDs the server named. Shared state then lives outside React, in a small store or the URL.

One screen doesn't need a rewrite

You don't need a front-end rewrite to make one screen feel like an app. A small render sidecar lets an existing back-end server render React components, hydrate only the islands that need to be interactive, and leave auth, the ORM, and your templates exactly where they are.

Worktrunk takes the friction out of Git worktrees

Git worktrees let you run agents in parallel, but every new worktree starts cold, fights over ports and leaves junk behind. Worktrunk wraps worktrees in a few simple commands and a hook system, so one config file gives the whole team warm, isolated environments that clean up after themselves.

Server actions aren't a React feature

Server actions look like React compiler magic, but underneath they are a name mapped to a function behind one endpoint. You can build the same shape in Django, call it from React or a plain HTML form, and keep validation at the boundary.

Thanks for all the buckets

MinIO deleted its community edition and pulled its Docker images, breaking every local and CI pipeline that used it to fake S3. Here is the four-step migration to RustFS on a Laravel stack, and why development dependencies deserve the same scrutiny as production ones.

A dial, not a switch: staging a risky migration with hash-bucket cohorts

A boolean feature flag starts a risky migration everywhere at once. Hashing a stable tenant ID into 100 buckets turns the flag into a dial that only ever widens, and the thresholds matter: percent-of-tenants is not percent-of-work.

The architecture diagram is dead

Infrastructure documentation goes stale because it is a second thing to maintain. Terraform plus cloud discovery plus a few query tools describes the system better than any diagram, and leaves humans to document only the why behind the odd decisions.

Subscribe