Automated anonymization
We have successfully reduced the amount of manual work on document anonymization using artificial intelligence and natural language processing methods.
RPA (Robotic Process Automation) as a technology is defined by one common factor - cost optimization: limiting the number of man-hours spent on performing relatively simple, repetitive tasks that can be defined in a rule-based manner.
As a company, we faced the task of automating the hiding of personal sensitive data in medical records, so that in the next steps such documents could be used without the possibility of identifying the patient. Based on the language models of deep machine learning and the regular expressions methodology, we have built a model that recognizes and then masks the above-mentioned sensitive personal data in documents.
Thanks to the implemented solutions, it was possible to fully automate this time-consuming process, saving up to 5-10 working minutes for each processed document.