Business Context
A crucial factor to the survival and evolution of organisations is to harness the power of Intelligent Automation.
Intelligent Automation has infinite potential in government, healthcare, retail and finance, specifically Machine Learning, which can discover patterns to make predictions. Chief Data Officers and Chief Information Officers can leverage this technology to predict healthcare demand and supply, determine deficiencies around service quality and provide predictive data on fraud.
Our client, International Personal Finance, harnessed the power of Machine Learning for fraud detection and their model now provides them with a clear data intelligence picture and predicts future fraud patterns, enabling them to stay ahead of the threat.
The Client
IPF is a leading international home credit and digital provider of consumer finance. They have 1.7m customers and have issued £772.2m in credit.
The Situation
- Legacy systems unable to optimise fraud analysis across 11 countries
- Early detection and analysis of millions of rows of data in seconds
- Pre-determined models that could not stay ahead of the curve and evolve to predict changing patterns in fraud behaviour
- The criticality of creating a sophisticated and optimal fraud detection system which itself acts as a deterrent for agents
The Solution
- Machine Learning implementation
- Detection of different types of fraudulent activity
- Business adoption within the SME community
- Correction of business model assumptions by co-locating SME and Mastek Analyst
"Mastek’s Machine Learning solution has been transformational—we have increased fraud detection by 15%."
– Amy Foster
Group Head of IT Change, International Personal Finance.
"Mastek’s Machine Learning solution has been transformational—we have increased fraud detection by 15%."
– Amy Foster
Group Head of IT Change, International Personal Finance.
The Solution
- Reduced the time taken to maintain the model from days to just minutes freeing up Analyst time
- Identified 80% of fraud cases
- Increased fraud detection by 15%
- Enabled 25% faster strategic global rollout
