Automating Tax Compliance Processes Using AI in under 30 days
The speed of results and very high levels of accuracy achieved by Kortical through the deployment of the Kortical platform were impressive.
Tax compliance processes are typically based on connected workflow tasks that require high levels of human effort for both data manipulation, processing and review. Tasks can be repetitive and time consuming to complete whilst others are complex, requiring subject matter expertise. Deploying automation techniques that feature optimal connectivity between machines and humans is challenging.
Kortical has demonstrated the potential to automate, with a very high level of confidence, key process tasks and is working with Deloitte on a range of projects focused on overcoming the challenges of compliance process automation.
Deloitte selected Kortical as its project partner following a competitive selection process because of the results achieved by Kortical over a 4 week testing period. For the specific task tested, accuracy and automation levels were in excess of 90%- a very strong performance over a relatively short training and testing period. Key to achieving these results is Kortical’s AI acceleration platform that uses AutoML to assist data-scientist in creating best in class machine learning solutions to a wide variety of problems- from predicating carbon emissions to automating email processing.
Automating compliance tasks requires layering many highly tuned machine learning models and Kortical facilitates this by dramatically reducing the model iteration process. The Kortical platform also hosts models in a production ready environment, automating the failover and redundancy process as well as automatically improving the models as they consume more data.
I spend a great deal of time working with our Ventures team in exploring the innovation space and have been impressed with how the Kortical team have applied themselves to the business challenges we posed them
Get the answers you need
Leave your email and we will contact you shortly or call us on +44(0)20 3998 3129.