Up to a 25% potential reduction in missed calls and 30% faster response times, indicating significant service
quality gains
AUTO APPROVE - AUTO LOAN REFINANCING
Revving up performance in auto loan refinancing
We worked with Auto Approve to analyze call center data using data science and AI techniques.
Our analysis uncovered key inefficiencies and highlighted opportunities to improve loan application completion, reduce missed calls, and enhance customer service through potential AI-driven optimizations.
About Auto Approve
Auto Approve is a leading provider of auto loan refinancing in the United States. Through its innovative platform, it enables customers to secure better interest rates and improve their financial health by restructuring their vehicle loans.
The company has served thousands of customers across the country, helping them optimize their budgets through tailored loan solutions.
The starting point
As Auto Approve’s primary channel for customer engagement, the Call Center was critical to business growth. However, the team faced major operational hurdles: nearly 1,000 unanswered calls on peak days, incomplete interactions, data inaccuracies, and high dropout rates. These inefficiencies translated into missed opportunities and limited scalability.
The company needed better visibility, smarter resource allocation, and tech-enabled workflows to improve performance and increase conversions.
THE GOAL
Unlocking efficiency with AI-powered insights
Our goal was to reduce call center inefficiencies and improve conversion rates by leveraging data science and AI to generate actionable business insights.
What we did
We partnered with Auto Approve to uncover operational inefficiencies and turn data into action. Starting with 16 business hypotheses, we focused on those with the highest potential for cost savings and performance gains.
Our team developed an exploratory, secure data pipeline to analyze patterns in agent behavior, call handling, and lead quality. We delivered predictive models and automated reports that highlighted optimization opportunities and supported smarter decision-making.
Key initiatives included:
- Discovery and prioritization of 16+ business questions and hypotheses
- Secure data access and analysis of behavioral patterns
- Development of validation models to uncover process inefficiencies
- Creation of a custom data pipeline to consolidate insights for analysis
- Setup of automated reports to guide ongoing decision-making
- Identification of staffing and workflow recommendations to potentially reduce call dropouts
A STRONG SOLUTION
The tech stack
The analysis leveraged a focused data and analytics toolkit, with Metabase providing intuitive dashboards to explore key metrics and test assumptions. Core Python libraries such as NumPy and Pandas enabled efficient data manipulation and processing, while Scikit-learn powered the machine learning models used to validate hypotheses and identify optimization opportunities.
This combination of tools supported the exploration of call center data, surfaced inefficiencies, and accelerated insight generation. By integrating advanced analytics in a controlled environment, Auto Approve gained a clearer picture of where data-driven improvements could be applied in the future.
Results
Smarter operations, better outcomes.
With our AI-powered data strategy, Auto Approve gained insights into how performance, efficiency, and customer experience could improve.
Models suggested a possible 20% increase in loan application completion, boosting conversion rates and revenue
Opportunities identified to streamline workflows through improved data access and potential automation of document handling
Analysis highlighted areas where customer satisfaction could be enhanced and friction in the refinancing
process reduced
- We are product builders.
- We are result accountable.
- We are strategic partners.
The Making Sense team members act as proactive collaborators who are empowered to contribute ideas, suggest improvements, and help build the best possible solutions—instead of simply acting as order takers who execute whatever is requested. They work alongside our employees, fully integrated into our processes and collaborating as one cohesive team.
Joe Traskos - Head of Product
at Auto Approve
at Auto Approve
Looking forward to making
sense with you
Tell us about your project challenges.