BGL says that its Simple Fund 360 AI Assistant is “the first Deep Learning transaction review assistant for the SMSF Software industry”.
“This is a huge achievement by our Simple Fund 360 Team,” said BGL Managing Director Ron Lesh.
“BGL has been working on AI for over 12 months. We needed to teach the AI Engine about SMSF transactions through a process known as Deep Learning. This results in Simple Fund 360 providing a level of confidence on where transactions have been allocated in the ledger based on the 250 million plus transactions in the Simple Fund 360 database. It is truly amazing technology,” he said.
“BGL’s AI Assistant can predict compliance breaches and spot incorrectly coded transactions that may have been input by a human or categorised in error by a rule.”
“The AI Assistant utilises purpose built Machine Learning classification algorithms to review bank transactions and attempts to locate any possible user or rule based coding errors and recommend coding corrections. This tool can be used, for example, during the pre-audit phase to locate outliers, improve processing consistency and reduce the time it currently takes to prepare a fund for Audit”.
“Deep Learning, an advanced form of Machine Learning, continuously looks for unusual patterns within transactions,” says
A recently released BGL video shows the AI Assistant in action. In one example a transaction is allocated as interest received, but the AI Assistant predicts – with 79% certainty – that it is in fact a dividend.
BGL says that the main objective of the AI Assistant is to reduce the time it takes to prepare an SMSF for audit, warning that it should be “used in conjunction with your normal review procedures”. The feature is also still in the beta phase of development.
“Stay tuned for Part 2 of the AI SMSF revolution, where further innovative and practical methods of Deep Leaning will be added to Simple Fund 360,” says BGL. in the video.
This article has been updated with comments by Ron Lesh.