ECOVIS Colombia has developed a machine learning model that identifies deviations or atypical behaviours in clients’ transactions. This implementation optimises the auditing and consulting processes and allows clients to react more quickly when anomalies are detected.
The analytics tools were developed by Ecovis to support auditing within the concept of data analysis using business intelligence (BI) models and the application of machine learning.
How the Project has Evolved During the Pandemic
Clients’ historical financial information was uploaded and at the same time an outlier, or anomaly detection model was developed through the construction of an algorithm using machine learning. The entire Ecovis team was trained on how to use the tool with clients.
These anomalies can be described as transactions that are outside of the normal behaviour and are identified after training the machine learning model using historical transactions. Anomalies or outliers should not be understood as errors and they should be validated with different audit tests.
The above implementation has been applied to everyday transactions and, more importantly, has enabled the Ecovis team to analyse what was almost impossible to inspect manually across the ocean of transactions reported by clients.
Such a detailed audit revision, using BI models, can provide insights into business performance and, alongside the risk matrix and internal control evaluation procedures, allows auditors to track business continuity milestones. Anomaly detection enables Ecovis to identify high impact transactions that may affect the financial statements to be reported, even if it turns out that they are within the normal process of the transaction, explain the Ecovis experts.
Ecovis has used the time during the COVID-19 crisis to develop the project on targeting audit practice. These models can certainly be used successfully in business consulting practice in post-pandemic times.
There is now greater visibility in front of clients and the opportunity to visualise risks in real time, say the Ecovis advisors.
For further information please contact: ECOVIS Colombia SAS, Bogota, Colombia