Site report extractor

The brief

Site reports arrive as free text, WhatsApp messages and photos of registers. Build an extractor that turns each report into structured records of activities, quantities, labour and materials, validates them, and measures accuracy. Good looks like: a schema, validation of units and quantities, and accuracy per field on 30 labelled reports.

What you will build

1. Write realistic site reports in several styles
2. Define the record schema
3. Extract with structured outputs and validate
4. Measure accuracy per field
5. Summarise progress from the extracted records

Free, open data to build it with

• eProcurement (CPPP): public works contracts ↗ Government of India

Check each dataset's licence before you publish your work. Your mentor gives your team its own version of the brief (data, scope or region), so every team's project is its own.