Etoundi Ondoua Ngando Emmanuel Tresor

Cadastra - AI-Powered Cadastral Forensics for Cameroon Land fraud and cadastral inconsistencies pose significant challenges to property transactions in Cameroon. One particularly concerning issue is "superficie fictive", where the area declared on a land plan may be significantly larger than the area enclosed by its actual boundary coordinates. Buyers and property professionals often lack accessible tools to independently identify these discrepancies before making important decisions. Cadastra addresses this problem by combining Meta AI technologies with geospatial engineering to automate cadastral document analysis and technical land-plan verification. Users can submit cadastral plans as PDFs or images. Meta AI assists in interpreting documents and extracting relevant information, such as boundary coordinates and cadastral markers. Cadastra then uses geospatial computation to calculate parcel areas, examine boundary geometry, identify coordinate inconsistencies and flag potential UTM zone errors. The platform also includes an AI-assisted analyst to help interpret findings, a WhatsApp-based verification workflow and bilingual PDF audit reports that organize the results for further examination. By combining AI-powered document understanding with deterministic geospatial verification, Cadastra goes beyond document scanning to provide actionable technical insights into potential land irregularities. The goal is to make preliminary cadastral verification more accessible, help people identify warning signs before property transactions, and contribute to greater transparency in Cameroon's land sector. Built for the Meta Global AI Developer Hackathon 2026.