Engineering
The engineering behind our products.
How SO HOMELY designs, builds and runs its own software: the principles we hold and the disciplines we invest in.
Engineering philosophy
Good software is designed to be understood, measured and maintained by the people who will own it for years.
Engineering principles
Clarity over cleverness
Readable code with clear boundaries outlives clever code that only its author understands.
Performance is a feature
Performance budgets are set before implementation and checked continuously, not audited afterwards.
Accessible by default
Interfaces are built to WCAG 2.2 AA and tested with a keyboard and assistive technology.
Proven foundations
We prefer mature, well-supported technology when a system has to run for years.
Six disciplines
The engineering disciplines every SO HOMELY product is built on.
Software architecture
We design systems around clear domain boundaries and explicit data models, and we write down the reasoning behind each decision.
- Boundaries drawn around the domain
- Typed contracts between components
- Decisions recorded in writing
AI & intelligent systems
We use artificial intelligence where it genuinely helps people work, and keep people in control of the decisions that matter.
- Practical automation and assistance
- Transparent, explainable behaviour
- Visibility into usage and cost
Platform engineering
Shared tooling, automated pipelines and consistent environments let each product move faster on a common foundation.
- Automated build and test pipelines
- Reproducible environments
- Reusable internal components
Performance
Performance budgets are set before implementation and measured on every change, so speed is designed in rather than tuned at the end.
- Budgets for loading and layout stability
- Static rendering wherever content allows
- As little client-side JavaScript as possible
Reliability
Systems are built to fail safely, recover quickly and stay observable when something goes wrong.
- Monitoring and alerting
- Backups with tested recovery
- Staged, reversible releases
Security
Security is a requirement from the first design review: least privilege, careful dependency management and secure defaults.
- Least-privilege access by default
- Dependency and vulnerability monitoring
- Content security policies and secure headers
Development principles
Understand
Map the problem, the constraints and the existing systems before writing code.
Design
Choose boundaries, data models and infrastructure deliberately, and write the reasoning down.
Build
Ship in small, reviewable increments behind automated tests and quality gates.
Operate
Measure real behaviour in production, close the gaps and keep documentation current.
Open source
Engineering work that SO HOMELY publishes under open-source licences.
QuotaOps
Open-source quota, capacity and local token-usage visibility for AI workflows.
- Type
- Open-source project
- Category
- Open source / AI tooling
- Platform
- Local web dashboard
- Licence
- Apache-2.0
- Repository
- github.com/ChrisLeoKokotos/quota-ops
View source on GitHub — QuotaOps (opens in a new tab)Read the case study — QuotaOps
Relevant technical work
How these principles shaped a real project.
Open source · 2026
QuotaOps: one view of AI capacity and usage
How it was built
- Local-first: no hosted backend, accounts or cloud sync
- Provider-neutral quota model with exact reset times
- Privacy by design: prompts and API keys stay out of its data
- Type
- Open source
- Year
- 2026
- Licence
- Apache-2.0
- Technology
- Next.js, React, TypeScript, Playwright
See the engineering in practice
Read how these principles shaped real projects.