Together with their integrator and consulting partner Codecentric, PLANET AI was presented with the challenge to revise and automate the digital mailbox for one of Europe’s largest telecommunications company. Their ratio of typewritten to handwritten documents is around 50:50 – proving handwriting as the biggest obstacle to overcome. The client’s former solution used Optical Character Recognition (OCR) and could not read handwriting from continuous text fast and accurately.
PLANET AI’s Solution:
PLANET AI managed to prevail against a direct competitor and provided Codecentric’s client with the IDA Textlayer capability. PLANET AI’s patented Automatic Text Recognition beat the previous approach by lengths and yielded significantly higher ratios regarding accuracy and speed. Hence, handwriting and typescript on scanned documents have now become accessible for reading and structuring. The client’s workflow is able to further process the Textlayer output without any adaptions necessary.
Together with Codecentric, PLANET AI implemented their IDA capabilities while minimally affecting and instantly improving the client’s current workflow. IDA Textlayer increased their automation ratio by around 25 percent – resulting in thousands of saved working hours and improved customer service responsiveness. Hard-to-read text, like handwritten IBANs, can now be extracted with high accuracy.
Other developments could include IDA Classification and Extraction to optimize the client’s input management system even further.