Klartext Extract
Vienna-based tool that turns long PDFs such as annual and sustainability reports into a fixed field schema, citing the source of every value.
What it does
Klartext Extract reads long PDFs in full, such as annual and sustainability reports, contracts or prospectuses, including scanned files with automatic text recognition. From them it fills a predefined schema of fields. Every value links to its page and passage and gets one of five confidence levels; conflicts between documents are flagged. Sums and ratios are calculated in code, not by the language model. In the review view you approve or correct values, and every change is logged. You can process whole folders as a batch and export results to Excel or JSON.
Who it suits
Teams in finance, credit assessment, rating, ESG and procurement that repeatedly pull the same data from extensive documents, for example coupon and maturity from bond prospectuses or notice periods from contracts.
Data protection for businesses
According to the vendor, hosting, storage and processing take place in the EU; enterprise customers get their own instance with a separate database. The terms state that a data processing agreement under Art. 28 GDPR is concluded. The vendor does not use customer data to train models offered to other customers unless you expressly consent. Third-party AI models and cloud services are used, which the vendor does not name.
Limits
Pricing is only available on request, and the vendor does not mention an API. You set up the schema together with the vendor, so you cannot start right away. You still need to review the values.