How we build, review and operate.
The standards behind the work — code, reviews, testing, architecture, documentation and operations. Each one backed by evidence you can read, none of them a mere claim.
Code
Business logic belongs in the domain, independent of the framework. We favor consistency over the locally best solution in each place, because it keeps the system readable as a whole — and we write for tomorrow's reader.
Evidence: Business logic doesn't belong in controllers →Reviews
Every recommendation states its price and the case in which we would decide differently. What we present as fact can be backed up; what is a position, we call a position.
Evidence: Engineering decisions (ADRs) →Testing
We test where failures are expensive — the business core and the paths whose breakage causes real damage. We prove the hardest path early and end to end, before building out the breadth.
Evidence: When Tests Really Pay Off →Architecture
The load-bearing decisions come first, made deliberately and documented. We start with the modular monolith and don't add structure until a concrete problem demands it.
Evidence: Modular Monolith vs. Microservices →Documentation
We think in writing. A decision without a recorded why is a risk: once the why is lost, every later change becomes a puzzle. We document what changes slowly — decisions, boundaries, operations — not what changes fast.
Evidence: Batunet Engineering Method →Operations
We design for failure and rehearse it before production does. Observability is part of day one; we ship in small, reversible steps and without downtime where availability matters.
Evidence: Zero-Downtime Database Migrations →