CarMax Skye+

Overview

Skye+ was a cross functional capstone project with CarMax focused on improving the car buying and financing experience for younger, budget conscious customers. While the project spanned product, marketing, and finance, I led product strategy and discovery.

Problem

Gen Z and younger buyers found the car buying process opaque, stressful, and financially intimidating. Interviews revealed confusion around:

• Tradeoffs between financing options
• Long term cost implications
• Trust in recommendations

CarMax wanted to explore whether an AI agent could guide customers through decisions without overwhelming them.

Discovery & Research

I led user research through empathy interviews and synthesized insights into personas and journey maps. A key insight was that users didn't want more information. They wanted confidence.

This shifted the solution away from static tools toward an interactive agent that could adapt recommendations based on user priorities and constraints.

Product Strategy

I applied dual track agile to separate discovery from delivery. My focus was on:

• Defining MVP scope
• Prioritizing features based on user value and feasibility
• Establishing success metrics around clarity and decision completion

I worked closely with teammates to align product decisions with marketing positioning and financial viability.

Solution

Skye+ was designed as an autonomous AI agent that:

• Collected user constraints and preferences
• Modeled tradeoffs across price, financing, and ownership costs
• Delivered explainable recommendations in plain language

I created low fidelity prototypes and experience flows in Miro and Figma to test assumptions before finalizing the concept.

Outcome & Learnings

The final concept was pitched to CarMax executives and received strong feedback around clarity and customer empathy. The biggest takeaway was how critical explainability is when AI intersects with financial decisions.

What I'd Do Next

• Pilot the agent with a limited customer segment
• Instrument decision points to measure drop off and confidence
• Explore deeper personalization through historical behavior