Case studies, white papers & news

Stay up to date about what is going on at BLP Digital and our highly innovative AI solutions.

Artificial Intelligence for information extraction from documents – ROI from day 1

Written on October 10th by BLP Digital AG.

The formula is simple. Reducing processing times + reducing error rates + increasing efficiency and productivity rates = increasing bottom line

At BLP Digital we work on a solution, which extracts all relevant information from business relevant documents using artificial intelligence. In order to better understand our formula in practice, let’s demonstrate the effects on invoice processing resp. accounts payable automation.

To manually process an invoice takes time. To thoroughly process an average complex invoice takes more than 3 minutes. Our solution significantly reduces this time, limiting it to reviewing and approving only. Further our algorithm works day and night for you, without getting tired but being concentrated 100% at all time. In combination with learning from uncertainties, corrections and changes, the Artificial Intelligence steadily increases accuracy and precision rates and is thus less prone to error than a human counterpart. The trick is to connect the two and let human operators constantly review uncertainties, that our solution highlights and correct them if necessary. That way the algorithm gets better and better and the operator can process a multiple of documents than before collaborating with our system. In our accounts payable example that leads to less operator time spend on processing the invoices, more invoices being payed on time, more exploited discount deductions and better supplier relationships, all of which result in an increased bottom line.

At BLP Digital we want to free up white collar employees from repetitive and tedious back office tasks by leveraging Artificial Intelligence. Our innovation is an automatic document processing solution, which extracts all relevant information from business relevant documents. We stand out with our solution as we recognize the complete document structure using a unique combination of deep learning algorithms for vision and natural language processing.

Contact us ›