Biotech company Centogene presented PLANET AI with the challenge to handle sample data with checkboxes, multiple forms and medical terms. Incoming samples were supposed to be classified and clinical data needed to be extracted and validated. Intelligent Text Recognition, especially for handwriting, had to be taken to another level. Their former Optical Character Recognition (OCR) could not fulfill the company’s needs regarding speed and accuracy – and therefore had to be replaced.
PLANET AI’s Solution:
- PLANET AI provided Centogene with their IDA engine (intelligent document analysis), including the SmartPDF feature, which unites all requested features. The textlayer contains the patented PerceptionMatrix, enabling content extraction and search. Additionally, a highlighting tool was implemented.
- The second feature – IDA Classification – only needs a small amount of training data to classify entire documents or single pages, which revolutionized sorting Centogene’s multi-page forms.
By implementing PLANET AI’s SmartPDF, Centogene gained a real AI assistant. Employees are now able to analyze and validate information directly in documents. Critical data, like checkboxes or keywords, can be marked in different colors and hence extracted.
The document classification assistant is easily trainable and adaptable to different languages. Thanks to IDA, the company can save thousands of working hours and speed up their workflow significantly. Additionally, on-premise dep