AI Takes the Wheel

In 2020, Carputty set out to disrupt the auto industry with an innovative line of credit. By 2021, rapid growth revealed a key challenge: a growing backlog of applications and increasingly dissatisfied customers.

Industry

Fintech

Role

Lead Product Designer

Designed

100% of Project

Team

  • Director of Product

  • 3 Lead Engineers

  • CTO

  • Startup Founders

The Vision

Traditional website templates often fail to effectively showcase photography and branding projects on larger formats, lacking a fullscreen horizontal experience that fully immerses the viewer in the content.

The Challange

Carputty's underwriting process relied heavily on manual workflows, compounded by technical debt and disparate system features. The result? Only 8% of applications were processed daily, leaving 92% of applicants waiting.

Research

I conducted 12 user interviews with applicants, Member Success Agents, and underwriters to identify pain points, which I mapped into a comprehensive user journey. The income verification step was responsible for 80% of delays, requiring time-intensive manual coordination (emails, calls, and resubmissions). Hotjar and Google Analytics confirmed this step as the primary drop-off point in the user journey.

Our Goals

Our primary goal was to reduce application wait times so we could grow our product and secure more venture funding.

  1. Underwriters should know how to use the platform day one

  2. Optimize and automate the credit approval process

  3. Implement real-time analytics to find friction

Key Decisions

  1. Familiar Interfaces: Developed an interface mirroring the underwriters’ manual spreadsheets to minimize the learning curve.

  2. Rich Data Display: Provided underwriters with comprehensive applicant details, improving their ability to make informed decisions.

  3. Real-Time Status Indicators: Designed clear status indicators to streamline collaboration between multiple underwriters working on the same applicant.

  4. AI-Driven Automation: I proposed and designed a system allowing underwriters to train AI directly through the interface. Over time, AI automated income data categorization has significantly reduced manual workload.

Significance

The new system transformed Carputty’s operations by addressing the bottleneck and creating lasting efficiencies:

  • 1200% increase in application throughput

  • 55 additional applications processed daily

  • $100M raised in funding shortly after implementation

  • 0 new underwriters hired, maximizing operational scalability

This project not only resolved critical bottlenecks but also became a cornerstone of Carputty's marketing and fundraising efforts. It exemplifies how thoughtful design and strategic automation drive impactful results.

Open to new work

Let's Chat About It.

Feel free to connect with me on LinkedIn.

Built in ATL

Andreas

Open to new work

Let's Chat About It.

Feel free to connect with me on LinkedIn.

Built in ATL

Andreas

Open to new work

Let's Chat About It.

Feel free to connect with me on LinkedIn.

Built in ATL

Andreas

Open to new work

Let's Chat About It.

Feel free to connect with me on LinkedIn.

Built in ATL

Andreas