AI helps, but not in longer run..
How AI ruined Projects & and time that I was invested, and later how I overcomed it.
Initial Architecture
In the early phase, the database schema was not fully optimized.
The relationships between:
- Test series
- Questions
- Attempts
were loosely defined.
Database was generated by AI, I ship a lot, not just feature but I were add lots of test series in db too. It was working smoothly for almost 6 months, then student came for different exam, they demand different type of test series, my db was not ready for this, hence, my complete backend code, db schema, types definition, ui component all suddently become irrelevant, and my project collapse.
Lessons ?
AI is a powerful assistant,
but system design decisions must be made by humans.
I should have to build system. which support flexibility, scalability, and maintainability. (I know these words might sound boring initially, but those who got taste, knows better.)
How did I overcome it?
(would be write after finishing my project, still I am working on it)