Structured Outputs & Schema-Driven Generation
Force the model to return valid, schema-conforming JSON every single time — because free text quietly breaks real applications.
What you'll learn
- Force the model to return valid, schema-conforming JSON every single time — because free text quietly breaks real applications.
- How structured outputs & schema-driven generation fits into the Advanced AI Engineering track
- A hands-on project step you can put in your portfolio
The hands-on project
Design a schema-driven extraction pipeline for parsing restaurant receipts into structured data. Write the JSON Schema (with field names, types, required fields, and at least one enum) for the fields a bookkeeping app would need: merchant, date, line items, subtotal, tax, and total. Then write the prompt you'd send alongside it, and describe your validation-and-repair loop: which business rules you'd check beyond types (e.g. do the line items sum to the subtotal?), and exactly what you'd send back to the model on a validation failure. Bonus: explain how you'd handle a receipt where the total is genuinely unreadable.
The full brief, voice tutor walkthrough, and feedback are inside the lesson.
Unlock Advanced AI Engineering with Builder
Builder unlocks all six tracks, unlimited voice tutoring, and the all-tracks Certificate of Achievement.