Senior Product Designer

Designing the decisions you don't see

Turning constraints into systems people can trust

01

Trust Infrastructure for a Two-Sided Marketplace

Established trust infrastructure across a marketplace of 400K+ agents and 150K+ loan officers.

Trust Systems Compliance as Product

When a platform acts on your behalf, failure isn’t UX. It’s liability. In a referral-driven marketplace, we were handling existing relationships and initiating cold outreach under strict RESPA constraints.

I turned compliance from a backend constraint into a user-facing system, absorbing regulatory complexity and translating it into distinct experiences for agents and loan officers so both sides could operate confidently within RESPA while the system enforced consistency at scale.

02

When Everything Is AI, Nothing Is

Standardized AI attribution across two platforms.

AI Product Strategy AI Governance

When everything is labeled AI, users stop trusting any of it. As Highway introduced Miles, an AI assistant and brand mascot, search, automation, and generative systems were collapsed under a single idea.

I reset how AI is represented across two products by defining a clear attribution model, separating deterministic systems from AI, and establishing standards for how AI is described and shipped as a product capability.

Getting attribution right forced a harder question: why use AI here at all?

03

Less AI, Doing More

Reduced AI system cost by 95% and made it viable for production use.

AI Systems Design Infrastructure

Highway’s design AI wasn’t scaling. At ~180,000 tokens per scan, a few runs could exhaust a weekly budget. I introduced a deterministic pre-analysis layer to control what reached the LLM, cutting token usage by 95%. From there, I restructured how the system made decisions—adding routing, orchestration, and evaluation layers to make outputs predictable and controllable.

This shifted the system from an expensive experiment to something that could scale reliably in production.