Advanced Analytics Framework for a Consumer Financing and Loan Servicing Company in the USA
To ensure financial institutions are on the right track, they must regularly measure their Key Performance Indicators (KPIs). For example, financial institutions could measure various parameters such as quality, cost of products offered, risk, or customer service. Moreover, by measuring this could ensure that the goals of financial institutions are completed smoothly in the long run.
Earlier KPIs were monitored using traditional Ad Hoc Analysis using markdowns and notebooks. Therefore, the data flew from hundreds of raw sources and had various cleaning and transformation that needed to be done. But, the entire process was manual, which was very time & resource consuming.
The goal was to transform and analyze the data seamlessly to gain actionable insights into the various metrics. So, the client wanted to have an advanced analytics framework with data lake integration, enhanced analytics, improved KPI reporting, and dashboard visualization. In addition, the entire pipeline needed to be completely automated to reduce manual efforts, and enhance monitoring of the Application Process, Approvals, Loan Originations, Loan Pool Performance & Other Initiatives. Therefore, this would improve and accelerate analytics-driven decisions and reduce the time for data analysis, data analytics, and data reporting on structured and unstructured data.
Data Lake
We built enterprise data lake architecture on the AWS platform. Additionally, this solution included AWS data lake architecture for scalable warehouse and AWS data lake architecture for structured and unstructured data. Also, the raw unstructured data would flow from various sources and get stored in the data lake, which would then be cleaned and processed for analytics.
Advanced Analytics
We wrote advanced analytics scripts to clean, preprocess, transform, and store unstructured raw data. In addition, we wrote scripts to calculate key performance indicators and other ad-hoc analytics of business use cases and demands. Hence, analytics and monitoring were done based on different segmentations, such as customer credit eligibility, pool vintages, etc.
KPI Dashboards
We built intuitive dashboards for visualizing key performance indicators and daily monitoring for stakeholders. Furthermore, the dashboard was divided into various tabs for different business areas. In addition, it had various control parameters to filter the dashboard across different segments, report dates, etc. Therefore, the dashboard involved monitoring the conversion funnel and understanding loan application trends, approval rates, and loan performance, among few.
Automation
The entire pipeline starting from procuring raw data from various sources to rendering intuitive visualizations was completely automated with different refresh schedules.
- Leadership Team at School
- Â Executives at Loan Servicing Team
- Student Advisors at School
- The visualization "Applications'' provides a 360-degree view of applications for student loans across various technical courses for different school campuses.
- The visualization "Approvals" establishes the approval rates for the loans applications along with monitoring of the UW Engine and ML Model.
- The visualization "Originations" provides information on the loan originated after getting approved and all its related metrics.
- The visualization "Loan Performance" identifies important KPIs related to performance of the loans after originations across different performing MOBs. In addition, it monitors the default rates and delinquencies in the loan pool.
All the dashboards had different control parameters to set various filters on the dashboard reports such as Date Filter, CScore Filter etc.
- The Aptus team began by examining the client's existing analytics architecture and methodology.
- Carried out a detailed requirement analysis and feasibility study.
- Presented the stakeholders with our updated advanced analytics framework with infrastructure on AWS along with scalability, performance & cost.
- Created ETL Pipeline to extract data from various sources and dump into data lake after cleaning, processing and transformation.
- Wrote optimized analytics scripts to utilize data and calculate key performance indicators and other ad hoc analysis.
- Built interactive dashboards visualizing the KPIs calculated in the previous step.
- Automated the entire pipeline & Tested the deliverables platform.
Python, R, AWS S3, AWS RDS, AWS EC2, AWS Lambda, AWS CloudWatch, AWS Quicksight, AWS Glue, Cron, PostGreSQL, Shell Scripting, DBeaver.
- The new advanced analytics framework benefited the client in multiple ways and helped to resolve the business challenge.
- Increased Productivity and reduced manual efforts largely.
- Better and enhanced visualization & monitoring of daily important KPI’s.
Related Case Studies
Achieving low-latency API-based queries with Mongo DB
Revolutionizing Pharma Analytics with AWS Data Lake
Boosting Performance with Apache Spark Migration
If you’re looking to take your business to the next level with data science, we invite you to contact us today to schedule a consultation. Our team will work with you to assess your current data landscape and develop a customized solution that will help you gain valuable insights and drive growth.