Design Leadership at Maven

Maven is the world's largest virtual clinic for women's and family health, serving members from preconception through menopause. When I became a design manager in October 2023, we'd just launched Maven Managed Benefit and earned our biggest client on the strength of it. The job was to prove it was worth what we'd promised, and keep that client.

Three years since launch, Maven Managed Benefit accounts for nearly $45M in annual revenue and close to 25% of enrolled members in Maven's fertility program.

Our original plan let members pay for treatment upfront and pursue care without a bill hanging over them, but that assumed a trust we hadn't earned yet with thousands of dollars at stake. We shifted to estimate upfront, collect after, which kept our transparency promise and made the estimate itself load-bearing. The client we launched with is still a client today, nearly three years later.

A year in, I fought to keep the cost estimate feature when operations wanted to cut it for accuracy and engineering-cost reasons. Estimates were a stated differentiator against competitors like Progyny, and members cited them in NPS feedback as what let them financially plan for treatment. We kept the feature and strengthened the qualifying language instead of rebuilding it outright. Over the next two and a half years, I led design to scale MMB through rapid growth.

Each of these started as a support ticket or a line in NPS feedback, not a design idea looking for a problem.

Care Delivery came under my scope in January 2025. Virtual care is Maven's core differentiator, with every provider touchpoint running through it.

MPractice has accumulated years of added data as Maven scaled, making it harder for providers to find what matters about a member. Rethinking its core architecture, unshipped, is my current focus, to keep it scalable as Maven becomes more clinically rigorous.

Any provider-facing feature a clinical program wants, like Maternity or Fertility, launches inside MPractice rather than as a separate surface. When a program needed providers to see a member's NICU status, that view went into MPractice too, keeping care longitudinal instead of siloed.

At launch, Maven Assistant deflected roughly 94% of member conversations without a human agent, over 99% for general health questions, and organizations with it enabled saw meaningfully higher engagement than those without.

Beyond the product, I led AI enablement for the design team as part of our VP of Design's broader push to make AI real across Maven, building the tools, knowledge, and workflows for the team to experiment with.

My Benefits background made fertility a natural fit: Maven Managed Benefit is a flagship product, but the fertility care program behind it is the differentiator.

Fertility's real goal is clinical: more members achieving pregnancy with no or low-intensity treatment. Since you can't design for a clinical outcome directly, the lever is engagement, getting members to actually use and return to the app so they get routed to the right care.

Every one of these efforts started with a question already costing someone something: a support ticket, a denied claim, a static dashboard nobody returned to, a provider typing through a session instead of being present in it. Benefits found those questions in what members were asking. Care Delivery found them in what was eating a provider's time. Fertility found them one layer removed, in the engagement standing between a good product and a clinical outcome. Assistant answered the same “is this covered” instinct with different technology.

None of that required a grand plan. My scope grew from one product area to four because each expansion followed proof that I could handle the one before it, and because the work kept making the case: revenue, retention, minutes saved, clients earned and retained.