Presolvent
Project Info
Project Description
Presolvent uses data from the Australian Financial Security Authority, in combination with ATO data to help predict when non-compliance might occur.
High-risk individuals identified by Presolvent can be given additional support and guidance about managing their finances, agreeing only to appropriate debt and insolvency agreements, so they can better meet their obligations.
Presolvent uses Machine Learning which is trained on limited datasets provided by the relevant Australian government bodies, it makes predictions of insolvency risk, but Presolvent is only a tool to be used in conjunction with human-assessed insolvency risk factors.
Data Story
We have combined a number of data sets into our application, including:
- The Australian Financial Security Authorities insolvency dataset that includes over 300,000 insolvency matters to train our machine learning prediction model
- ATO Tax return data set to normalise distinguish individual types from insolvent individual profile types.
Team DataSets
Insolvency and Trustee Service Australia FOI Disclosure Log
Description of Use: For more data points into the insolvency prediction model.
ATO Dataset: Gov Hack 2018
Description of Use: We used to profile individual types that are solvent, compared to individual types that are insolvent in AFSA's dataset
Non-compliance in personal insolvencies
Description of Use: We used to train our machine learning prediction model which assesses risk of non-compliance with debt agreement.
Challenges
The Friendly ATO
Region: Australia
ChallengeTo bankruptcy or not to bankruptcy, keeping the process real.
Region: Australia
ChallengeData4Good
Region: New South Wales
ChallengeChatbots are the Future
Region: New South Wales
ChallengeShow Us The Numbers
Region: Australia
Challenge