Handling activities for this many properties involves hundreds of staff and thousands of contractors. Most of the focus is put on handling the requests in a timely fashion and ensuring that the work performed is of high quality. The back-end filing of the data becomes a secondary priority. The importance of filing is critical however as it measures whether both client and legal obligations are met.
The challenge for the business was how do they quickly and easily categorise all the recently completed requests? It would take a human approximately one month of tedious work to get through just a fraction of the workload. The answer was in the data itself. With the use of AI Computer Merchants developed a machine learning model to perform the categorisation. In less than 1 day machine learning was able to categorise over 18,000 documents and return a compatible data set to be integrated into their corporate systems. Importantly, the accuracy was comparable to what you would expect from a trained employee.
The next step for the organisation is to integrate the AI model into their daily workflow so that the document categorisation happens in real time. The IT team have also identified their next use case for AI is to help manage and detect fraudulent invoices.