Healthcare Technology Use Case


Large hospital systems send our client hundreds of thousands of dynamically unstructured healthcare documents a day. These documents need to be classified and depending on the document type, certain information needs to be extracted and populated into each client’s electronic health record system (EHR). For years, this was a manual process that was outsourced to humans overseas to read the documents, manually classify and type out the required data points into an Excel spreadsheet. The issue with manual extraction is that it is error-prone (humans average error rate is 10-30%), extremely time consumed, and thus costly. Additionally, sensitive healthcare information is being shared which created unnecessary security risks and/or potential HIPAA violations.



Large healthcare data aggregator                                                                                                                                                                                                                                                                                                           

Tackle Ai's Solution:

The old workflow described above used to take up to 48 hours. After leveraging TackleAI, this same workflow is now intelligently completed in fractions of a second. Documents such as prescriptions, MRIs, oncology reports, etc. are all dynamically processed through the TackleAI engine, which extracts the relevant data points requested by our client. Data points such as patient name, physician name, date of service, medical record number, social security number, addresses, medications, and readings, etc. Most documents the TackleAI engine is processing have never been seen by the AI before.


TackleAI is faster, more accurate, and 70% less expensive than human labor overseas. We are saving our client 50% of their overall operational budget. 73% of this customer’s documents are now being completely processed with no human intervention. Human error rates were 7.5% with the old manual workflow, but after the TackleAI engine was introduced, this error rate has been dramatically reduced to 0.03%