How we use AI like engineers
Using AI to recover context in large codebases and take repetitive work off engineers' plates without compromising on code and review quality or operational safety.
Blog
Engineering writing
Selected reposts and engineering writing relevant to Diversio Engineering.
No Code by Hand
CI dropped from 37 minutes to 9, at 35% lower cost. What Diversio learned about mixing Claude and Codex, clearing the hidden queue, and where the real multiplier came from.
Read article →Protecting Our Own Tenant in a Multi-Tenant SaaS
How Diversio treats its own tenant inside Optimo as the hardest one to reach, using layered controls across Django admin, approvals, Postgres RLS, IAM, and automation.
Read article →The Monolith That Made AI Actually Useful
How Diversio solved context switching across multiple repositories by building a monolith using git submodules, making both humans and AI 10x more effective with the codebase.
Read article →From Postman to Bruno - How AI Changed Our API Workflow
Diversio migrated the entire Postman collection to Bruno over a weekend and leveraged Claude Code to automate API documentation, reducing documentation time by 90% and catching breaking changes at review time.
Read article →Managing Context Metadata in Django-PGHistory
How Diversio solved metadata leakage between contexts in django-pghistory for cleaner, more accurate audit trails, and what other Django teams can learn from the approach.
Read article →Evolving Engineering Recruitment at Diversio in the Age of AI
What Diversio learned about hiring engineers as the landscape shifted from take-home exercises and resume screening to AI-assisted filtering and streamlined interviews.
Read article →