OCR: scans become data
The flinq OCR is in development: domain-specific text recognition for construction documents. It will turn scans and photos into machine-readable text and structure.
What is OCR?
OCR stands for optical character recognition. An OCR model reads documents that only exist as images, such as scans or photos, and converts them into machine-readable text. Good OCR also recognises the structure: headings, tables, line items and quantities are preserved as such.
Only this step makes a document usable for software. What was a pixel image before becomes text and structure that can be searched, extracted and processed further.
What does that mean for construction?
A large part of the knowledge in construction sits in documents that were never digitally structured: scanned tenders, older bills of quantities that only exist as PDF, plans and product datasheets. Generic text recognition often fails on the terminology, short texts, units and table layouts of these documents.
The flinq OCR will be built for construction documents and make these archives readable. The result feeds directly into the other flinq building blocks: search and matching via the Embedding and Reranking API.
The OCR is in development. Subscribe to the newsletter so you do not miss the launch, or write to us.