Leading a greenfield SaaS for technical debt

##The problem

Every organisation knows it has technical debt. Few can say where it is, how bad it is, or which part to fix first. Code Tuner measures signals like complexity, coupling and how often code changes, without running or changing the code, and presents them so engineering and business can prioritise together.

##What I did

• Architecture

  Designed and built the microservices and REST APIs with NestJS and Hono on PostgreSQL via Prisma, with Python where the analysis needed it.

• Front-end

  Built the React and Next.js application, including data visualisations with D3.js and Chart.js that make metrics readable for people who do not read code.

• Quality

  Set up testing and CI/CD with Jest, Vitest, React Testing Library and GitHub Actions, and used Stryker mutation testing to check that the tests actually catch bugs.

• Leadership

  Led and mentored the team, and ran the Scrum ceremonies as Scrum Master while working closely with the Product Owner.

##Decisions and trade-offs

A tool that measures technical debt cannot have a lot of its own. That is why testing and CI/CD were part of the first sprints, not something to add later, and why mutation testing was worth its extra run time.

Being lead developer and Scrum Master at once is a balance. It helped to keep the two roles visibly separate: in refinement I spoke for the technical side, in the retrospective I facilitated and let the team talk.

##Result

Code Tuner went from a greenfield idea to a working SaaS platform that is now live at codetuner.digital.