Real Estate: Automatic extraction of thousands of PDF invoices

Context

A real estate operator specializing in offices, meeting centers and coworking spaces wanted to automate the processing of its Thousands of invoices, in order to save time and improve energy monitoring at the scale of several buildings.

Challenge

• Key information (volumes, counters) not extracted from invoice PDFs.
• Impossibility of cross-referencing energy data with accounting data.
• Time-consuming manual classification of invoices by type (energy, maintenance, etc.).

StratBoard Solution

• AI test on automatic extraction of consumption data from PDFs (electricity, gas, water).
• Creation of a multi-site energy dashboard, centralizing consumption per building and per meter.
• Phase 2 : automatic categorization of invoices by type via OCR and LLM.

Results

✅ Reliable treatment of Thousands of invoices PDF.
✅ Centralized energy monitoring on 2 pilot sites, with visualization of trends.
✅ Automatic pre-classification bills (energy, maintenance, elevators...).

Conclusion

StratBoard proves the effectiveness of AI to automate the processing of invoices in real estate on a large scale, by facilitating the analysis, categorization and management of consumption.