Skip to main content


Over one million letters per year are sent to this government agency and each of them needs to be processed before a decision can be made and a response can be sent. The work preparation steps were done manually which often resulted in capacity challenges, SLA exceeding lead times and additional work due to human errors that needed to be corrected downstream.


We developed a platform that interprets, extracts and validates all relevant information needed to prepare a case for decision-making. The platform was made generic enabling it to be rolled out into other similar processes within the organisation. It allowed the business to continuously monitor and improve the solution and the quality parameters to the best trade-off between automation and downstream error handling.


The project resulted in the automation of 50% of the work preparation for the initial process, reduction of case file anonymisation lead times from 30 to 3 days and reduced response times from 3 months to 3 weeks for high-volume cases through batch-based processing.