Project Info

Hard Pivot thumbnail

Team Name

Hard Pivot

Team Members

8 members with unpublished profiles.

Project Description

Problem Statement

Every year more than 30,000 Australians become insolvent, owing creditors more than 7 billion dollars. Insolvency grew 7% in the 2017-18 financial year - double the rate of bankruptcies.

The reasons for insolvency are complex and varied, but impact everyone in our community. Total bankruptcy costs the Australian economy 0.5% GDP per annum.

About 1,000 people per year fail to comply with their insolvency agreements, causing more distress to creditors and costing AFSA millions of dollars.

Introducing INSOLVED

Insolved is an advanced analytics engine to help everyday Australians get back on track.

Using machine learning and predictive modelling, Insolved estimates with 95% accuracy if people within the personal insolvency system are at a high risk of non-compliance (for this project, we define non-compliance as Offence Referral and Objection to Discharge).

With Insolved AFSA can:

  • Predict non-compliance with 95% accuracy to take appropriate measures
  • Understand trends in insolvency and non-compliance as well as profiles of the average Australian facing insolvency
  • Understand regions of financial stress in Australia and glean insights into root causes so that early intervention or financial education drives can be mobilised
  • Understand trends for optimal utilisation of resources for the benefit of trustees and individuals

Here's how it works:

- An AFSA employee will have a dashboard of all personal insolvency cases.
- Each case shows a risk rating of High Medium or Low, based on our algorithm.
- The case list can be sorted by Risk Rating, ID number or Year and filters can be applied to highlight different cases
- Alternatively, employees can find a specific case by using the ID search

Prediction tool
- To predict whether someone within the personal insolvency system is at risk of non-compliance, all you have to do is fill in a few details.
- Our predictive model will determine whether this person is low, medium or high risk.

We’ve compared the personal insolvency data to a variety of other datasets looking at health, housing, education and employment across Australia.

We’ve found several interesting trends which can be explored within our Tableau model and are visualised in our product
We found that in the top 10 geographical regions with high rates of insolvency:

  • Unemployment rates were 12% higher
  • Psychological distress rates were 17% higher
  • Obesity rates were 20% higher
  • House and unit price growth were 74% and 119% higher
  • The number of people with a bachelor's degree was 28% lower

... when compared to the national average.


Insolved has the potential to drive down the cost of administration of non-compliance and reduce non-compliant debt.

But we can go further:

  • Insolved could be used by the public to predict their risk of becoming insolvent and receive advice on reducing financial stress.

  • This information could further enable AFSA to predict trends in insolvency by location, profession, gender and family circumstances, better preparing them to deploy resources and launch education campaigns.

Data Story

Our predictive model is unique

Our algorithm solves the class imbalanced or needle in haystack problem as the number of those who are non-compliant is small. We needed to come up with accurate predictions when there was an imbalanced distribution of non-compliance data within a highly skewed dataset.

Novel and creative data insights

We developed ways to correlate insolvency data and show that insolvency affects us all Australians but often affects some of the least fortunate. We did this by combining health, education, economic and housing affordability data. We also developed profiles of male and female insolvents to show insolvency affects average Australians. Check out the models/text document as we were not able to visualise everything. Data preview below


Evidence of Work



Team DataSets

5 years of ABS Regional Data

Description of Use Correlating health, education, housing affordability and employment to regions of high insolvency

Data Set

ABS social economic equality index

Description of Use Relative advantage and disadvantage of socio-economic circumstances correlated to insolvency

Data Set

Australian Health Tracker

Description of Use Correlation of insolvency to chronic lifestyle diseases

Data Set

Non-compliance in personal insolvencies

Data Set

Challenge Entries

To bankruptcy or not to bankruptcy, keeping the process real.

Helping predict non-compliance in the personal insolvency system. How can Artificial Intelligence and Machine Learning assist us in the future?

Go to Challenge | 13 teams have entered this challenge.


How can open data be used to make a social impact, contributing to the betterment of society? How can we improve prospects for children, and education, using open data? What sort of impact can be made on homelessness, mental health outcomes, or the environment, using open data?

Go to Challenge | 19 teams have entered this challenge.

Spatial data challenge

How can spatial data be leveraged to provide the best community outcome? How can this mapping data be used to deliver value to the people of NSW?

Go to Challenge | 14 teams have entered this challenge.

More than apps and maps: help government decide with data

How can we combine data to help government make their big and small decisions? Government makes decisions every day—with long term consequences such as the location of a school, or on a small scale such as the rostering of helpdesk staff.

Go to Challenge | 58 teams have entered this challenge.

Bounty: Is seeing truely believing?

How can we tell a story with visualisations, that speaks the truest representation of our data?

Go to Challenge | 28 teams have entered this challenge.