TM
Back to Portfolio
Consumer AI & Personalisation Technology

Project Mana Menu

Enterprise API Discipline Applied to Consumer Product

~657
Automated tests
37
Controllers
554
Quality-gated recipes

The Challenge

Building a meal-planning backend that could personalise weekly menus across allergens, cuisines, schedules, and portion models — while supporting a bilingual recipe catalogue and quality-gated content pipeline. Cooperation project.

What I Built / Delivered

Built a .NET 9 API with Clean Architecture + CQRS: ~107 MediatR handlers, 37 controllers, 70+ domain entities. Pluggable menu-generation algorithms with multi-dimensional constraint-satisfaction scoring. ELT recipe pipeline with quality-gated promotion. JWT auth with refresh token rotation, Google OAuth. Full Docker/nginx deployment with CI/CD.

The Outcome

~657 automated tests with CI-enforced controller coverage. 554 recipes promoted through quality gates. Production readiness checklist with all P0/P1 items complete. Multi-algorithm personalisation with documented satisfaction scoring.

Why This Matters

Enterprise API discipline applied to a consumer product — Clean Architecture, CQRS, and rigorous testing are not overhead but the foundation of reliable delivery at any scale.