Classification Engine is commonly used where a human has large volumes of data that requires a manual line-by-line review. Classification Engine's machine learning technology will learn from previously reviewed data and use this knowledge to make future decisions. PwC has helped companies save time on a range of tasks that require manual analysis, to ultimately increase integrity and reduce burdensome tasks.
Classification Engine automates the line-by-line review of your entertainment expenditure to determine the appropriate treatment of each expense for Fringe Benefits Tax (FBT) purposes. Classification Engine increases the integrity of your review process and identifies opportunities to ensure the most optimal treatment or methodology is selected to minimise your FBT liability.
Playback of this video is not currently available
Classification Engine is frequently used to analyse project (WIP/AUC) expenditure and Fixed Asset Registers. Its primary use is to highlight areas of risk or opportunity by determining the correct treatment of each expense item (e.g.deductible itself, treated as a separate asset or form part of a project cost base). Classification Engine can also be used to process prior year data to locate errors and identify potential deductible transactions to submit an amended tax return.
Classification Engine is an extremely flexible classification tool. Anytime your organisation needs to classify large data sets into buckets, Classification Engine can automate this process. Our clients are using Classification Engine to classify expenditure into accounts for accounting purposes, organise procurement data in order to understand volumes of purchasing and classify payments to different categories of vendors for various reporting requirements. Classification Engine can be set up in minutes and even assist with one-off jobs. Classification Engine will tirelessly and faithfully replicate the classification decisions of your skilled employees over large data sets its never seen before, freeing them up to focus on value-accretive tasks. Get in touch to explore the variety of use cases for Classification Engine.
"Prior to Classification Engine, all our FBT workpapers have been prepared in excel and manually reviewed. We spent weeks analysing our general ledger, line by line. Due to the large volume of transaction in our business (across multiple entities), human errors may have occurred and transactions may have been misclassified in the past. With the implementation of Classification Engine, we have seen an improvement in efficiency as well as savings in terms of actual FBT liability. Weeks of analysing data has been reduced to days. Our internal processes have been enhanced as we are now analysing data on a regular basis and therefore taking advantage of the machine learning capability that the Classification Engine technology provides. We are also generating monthly FBT reports which make the end of FBT year processes easier to provide further insights into our FBT spending."
Teach Classification Engine by uploading prior year workpapers or classifying raw data "in system". Then based on previously learned data, Classification Engine will receive raw datasets and automatically perform the line-by-line analysis.
On occasions that Classification Engine is unable to make a determination, it uses its growing intelligence to make suggestions and allow the user to review and manually classify the data directly within Classification Engine.
Users can review Classification Engine's decisions, make changes and provide feedback to Swift Classifier for future decision making. These changes contribute to the continuous education of Classification Engine's machine learning capabilities.
With the elimination of line-by-line review, teams are empowered with the time to review the results and make important business decisions based on the output. Classification Engine is equipped with integrated analytics to visually review the data, export to Excel or integrate with ERP systems for end-to-end automation.