Enter the password to view case study 👀
AI MESSAGING CO-PILOT
September 2025 – January 2026
Designing a 0-to-1 context-aware AI recommendation engine that analyzes buyer intent to accelerate dealer responses and drive carbuyers down the sales funnel
project CONTEXT
This feature is part of a growing AI capabilities in Dealer Navigator product team. This is the second AI capability that streamlines a dealer's workflow. Dealer messages tend to be templated scripts based on what stage the customer is at.
This tool empowers dealers to send the right message to the carbuyer at any stage in the carbuying process. In addition, we heard in dealer research that some salespeople don't speak English as their first language or have less schooling, so this helps create professional, tailored messages.
My role
Lead UX Designer — owned dealer research, product design, and design QA partnership with engineering through to launch.
Snapshot
AI/UX Design · Research synthesis, rubric design, cross-functional design QA · Rubric now governs 2 shipped AI products
Discover
Ran a competitive analysis across six competitors, gathering screenshots and demo footage from public-facing materials. For the first round of dealer research, dealers were asked to react to 3 different concepts and annotate their own ideal solution on paper — a generative method that surfaced assumptions a typical concept test would have missed.
Define
Contributed to a single AI rubric spanning two products (AI Recommended Messages and Chat Concierge, a chatbot that lives on a dealership's website). Assessed which buying intents were in scope (e.g., scheduling an appointment) versus out of scope (e.g., estimated financing terms), then defined quality dimensions (message quality) and hard safety guardrails (message monitoring).
Develop
Ran a second round of dealer research to pressure-test interactions and expected AI behaviors, like follow-up messaging. Used findings to redesign the feature's entry point and identify the need for a larger dedicated messaging surface — validated at a cross-team design review before moving to MVP.
Deliver
Shipped the MVP in November 2025, staying embedded with another team’s tech team while the backend AI model was built in parallel by my team. Provided ongoing async design QA — screen recording walkthroughs for a developer on a different team and annotating screenshots — and sat on a rotating cross-functional pod that audited live AI-generated messages against the rubric to ensure the messages were within scope and identifying the right intent.
Impact
This AI feature empowers dealers to instantly respond to customer inquiries with highly accurate, compliant, and context-aware suggestions. For car buyers, it delivers the immediate, personalized answers they need, progressing them down the sales funnel.
One rubric now governs message quality and safety across two shipped AI products, reducing redundant compliance review and giving both teams a shared definition of “a good AI message.”
"As a product manager, I had the opportunity to work with Linh-Hoa to build GenAI features for one of our dealer-facing products at Capital One Auto Finance.
Linh-Hoa led our consumer research and synthesis, ensuring our user experience directly addressed genuine customer needs. She then transformed those insights into a seamless, intuitive product design. With an inquisitive mindset and strong execution skills, Linh-Hoa was a great partner to collaborate with and also truly brought joy to the team!"
-Chris Spahis, Senior Manager, Digital Product Management at Capital One
"Linh-Hoa is an exceptional product designer who brings a great balance of problem-solving and strong execution to a team. We worked together on the Recommended Messages feature, where she took a complex interaction model and turned it into an elegant, clear user experience.
She excels at translating abstract product ideas into practical designs. Her communication is straightforward and collaborative, making it easy for her to align cross-functional teams and move projects forward smoothly, all while delivering customer value quickly. It was a pleasure working with her!"
-Sophie Babo, Manager, Product Management at Capital One
