How VetOps was discovered through ChatGPT
VetOps getting discovered through ChatGPT changed how I think about product positioning: language now has to serve humans, search engines, and AI assistants at the same time.
NOTES
Notes from building software — architecture, product decisions, and lessons from production systems.
VetOps getting discovered through ChatGPT changed how I think about product positioning: language now has to serve humans, search engines, and AI assistants at the same time.
PATTY CLI started as 50 lines of shell script. Two years later, it handles deploys, backups, rollbacks, and server diagnostics across multiple projects. The lesson: start ugly, ship fast, let real needs drive the roadmap.
Offline-first is not a feature you bolt on. It is an architectural constraint that reshapes how you model data, handle conflicts, and think about reliability.
The gap between how work happens and how software models it is where most operational tools break. Closing that gap means understanding the job before writing a single line of code.
A 50% performance improvement and 20% infrastructure cost reduction did not come from switching frameworks. It came from rethinking the data layer, query patterns, and deployment architecture.
Most feature requests describe a symptom, not a solution. Learning to translate requests into architectural decisions is what separates software that survives from software that has to be rewritten.