Credit Risk Scoring Engine
- Table of Content
- Company profile where Muoro helped with advanced algorithm systems
- Goal of the Company:
- How Muoro helped them with advanced statistical algorithms?
- Which technologies were adopted by the experts?
- What did they achieve?
A leading credit management group specializing in debt collection, business information & credit reports and Para-legal services.
Provide a fair and accurate user profile and maintain a consistent record.
Why company needed advanced programming solutions?
The client wanted to enhance the current credit risk methodology and enable the learning algorithms to increase the efficiency of current business reports. Furthermore, they need a better data analytics suite from basic excel modeling to advanced statistics that can provide them good visualization.
Even after we had some pre-defined advanced statistical models with us, we developed the advanced learning algorithms based on the client’s input. This smart analytical system has customized algorithms relevant to the client’s needs and industry.
Provided the system to ease the analytics job for their analysts with a complete dashboard offering collaborative drag &drop features with Tableau.
The system was designed using the technologies as followed:
Database – NoSQL(MongoDB)
Frontend – Integration with Tableau for Visualizations
Backend – Node.js
Analytics and statistical models – Used Python and associated libraries (NumPy, pandas and PyTorch)
API – Developed REST APIs for internal connections with .NET Framework
The new predictive credit risk models helped the company to evolve their obsolete methodology, the decision tree analysis included a new set of parameters that were segregated according to the learning algorithm, hence automating the task and increasing the efficiency:
- 9% increase in revenue through new business reports.
- 12% decrease in the total time cycle to generate the business reports.
- 89% Accuracy rate of the scoring engine.
