Tolstoy’s Data Extractor extracts and labels text from virtually any document or record.

Unlike other solutions, we customize an Extractor for each client. Because of this, we are able to process highly unstructured documents — from legal transcripts, to email data, to handwritten records.

The university digitized a collection of one million documents from the Nuremberg War Crimes Trial. The contents of the entire collection was unknown, so the university partnered with Tolstoy to help index the documents. This was done by identifying all document references in their 150,000+ pages of court transcripts.

Tolstoy tagged the references with 99%+ accuracy. This saved them approx. 7-8 months of work.

To celebrate their 130th anniversary, The Wall Street Journal wanted to digitize and reprint articles from their entire history in a special edition. Since many of the articles were very old, traditional OCR software did not pick up the text well or at all.

We wrote a custom OCR script that parsed their digitized articles with 95%+ accuracy. This saved them several weeks of manual transcription.

The New York Power Authority (NYPA) has a critical, 660-megawatt power cable that experienced multiple faults, taking it out of service for 1 to 2 months at a time.

With only disconnected paper documents, NYPA needed one holistic and connected view of the data to properly assess the situation.

Using a customized AI text extraction model, Tolstoy digitized the documents into an Excel database that allowed NYPA to easily view and evaluate the information to determine critical next steps.

Museums across the world are digitizing their records. This includes extracting text from specimen labels into databases. The labels are often handwritten, with an unknown number of formats. The labels come from various eras, countries, and institutions.

Tolstoy built an Extractor that processes print labels with 98%+ accuracy, and handwritten labels with ~80% accuracy, regardless of the format. This project is ongoing and has the potential to reduce processing times from decades to weeks.