When We Stopped Using Microservices (And When We Kept Them)
A modular monolith is the right default. Microservices win in two specific cases. Here is the rule we use and the projects where we changed our mind.
in Engineering
— /journal
Engineering notes & field stories
Production-tested patterns, product stories from Blink AI, Nexia Academy and Wallnetic, and the trade-offs we hit shipping real software. Written by senior engineers, not marketing.
-
-
B2B SaaS Onboarding: An Email-First Pattern Library
Why email beats in-app for B2B SaaS onboarding, the seven email patterns we ship for new tenants, and what each one is actually for.
in Engineering -
Beyond GROW: Other Coaching Methodologies for AI Products
CLEAR, OSKAR, IGROW and others — coaching methodologies that map well to AI product design, and the trade-offs of each compared to GROW.
in AI Products -
Cost Per Coaching Session: How We Track AI Unit Economics
How we instrument every AI coaching session to track cost per session — and why this is the metric that drives every other AI product decision.
-
Multi-Region Postgres on Hetzner: A Practical Guide
Setting up multi-region Postgres on Hetzner for SaaS read scaling and DR — async replication, failover orchestration, the observability that makes it operable.
-
Anatomy of a Production Incident: Our Postmortem Template
A blameless postmortem template, the seven sections we always fill in, and why “what would have caught this earlier?” is the question most teams skip.
-
How We Hire Senior Engineers (Without a Brain Teaser)
Our hiring loop replaces algorithm puzzles with paid work-sample projects, real code review and architecture discussion — and the patterns we use to judge senior engineering ability.
-
Offline-First Mobile Apps: Architecture Patterns
Local-first data, sync queues, conflict resolution and optimistic UI — the architecture patterns that make offline-first apps feel native, not aspirational.
in Engineering -
Cross-Platform App Performance: Where Time Actually Goes
Profiling shows time goes to four places: bridge cost, render commits, native module init and image decoding. Where to optimize each and where the wins actually are.
in Engineering -
LLM Cost Optimization: Cutting Tokens Without Cutting Quality
Six techniques that cut our LLM bill 40-60% without dropping coaching quality — prompt caching, smaller routing models, summarization, structured outputs and more.