Design decisions
Five decisions, two minutes each
Each page answers one question: what was decided, what it costs, and what was shown.
- The model plans; the pipeline writes the SQLWhy doesn't the AI write the query?
- Checks decide what gets publishedHow does a query that runs still get rejected?
- Spend is reserved before a run startsHow can a chat tool have a hard budget?
- The backend authors the screenWhy can a reader trust what the page shows?
- Evaluation is sealed, and a failure countsHow do you know it's accurate, and how accurate?