Teams are systems too. They have interfaces, feedback loops, failure modes, and moments of magic. I have been designing them since 2014, and I am still not done learning how.
I started leading at a company I co-founded. Royal Cause was a small venture with a team of designers, researchers, and marketers, and it taught me the whole job at once: hiring, coaching, shipping, and answering for all three. At Oracle I led a design workstream inside a very large machine, and with the research team built a self serve program that taught designers to run their own studies: what to do at each stage, what to know, and when to call in a researcher. At Vendia I ran design across product and marketing, evolving a blended agency and in-house org into a lean senior in-house team. And at MongoDB I lead UX Foundations, the org behind Via.
The first job of leading a team is telling them where we are going, twice. There is the north star, far enough out to organize years of work, and the near term, the next six to twelve months, close enough to act on this week. I hold both to the same rule: a goal is a thing someone can do, not a description of how we will do it. Via for anyone at MongoDB is a goal. So is generating working product experiences that are right, or close to right, on the first try. When the destination is stated as an ability, every team member can invent their own path to it, and the best paths are usually ones I would never have assigned.
The part of the job I take most seriously is who the team becomes, and that starts with actually knowing the team. When I inherited UX Foundations I talked with everyone, reports and skip reports alike, until I had a real picture of what was working and what was not. Trust like that is earned by what you do with it. Sometimes acting on what I heard meant changing something. Sometimes it meant digging in and coming back with an honest, unbiased answer. Either way, people learned that telling me the truth led somewhere, and psychological safety rose with a sense that we were in it together. So when leadership turnover came, some of it necessary, the trust held. I brought in proven leads, and the ICs who would report to them believed I was choosing the right people, because I had earned the benefit of the doubt. I sponsor promotion cases with the same rigor I bring to design reviews, because a growth story deserves evidence.
I rebuilt how my team works around a simple bet: AI should absorb the repetitive judgment, so the team's time goes where judgment is scarce. Inside the team, designers and engineers use it to build and improve Via and the quality layer behind it. What we ship extends that bet across MongoDB, giving every designer the same trade: less time on the repetitive, more for the new, the ambiguous, and the hard.
We made learning structural. AI Fridays gave everyone protected time to play: small, controlled experiments with new tools to work out what belongs in daily practice. When MongoDB ran its company innovation week, my team participated at 100%, with the next closest team in our org at 62%. We share what we learn on a biweekly rhythm, and every six weeks we review the work itself, asking which tasks have earned their way into shared skills and evals. The result is a team that spends less time proving quality and more time raising it.
Recursive Thinking started in 2015 with a gap my husband and I could not leave alone. People around us wanted to learn to code and had quietly decided it was too hard for them. We disagreed, so we started hosting Saturday practice sessions in our living room. Then people asked to bring friends, and the friends brought friends, and one Saturday we counted fifty people in a tiny Seattle house, spilling into the dining room, the backyard, and the stairwell, which it turns out has an excellent view of the TV. The stairwell was the moment we admitted this needed real structure: group sessions, a rhythm people could count on, and eventually a name. When the pandemic took the living room away, it moved to scheduled sessions on Zoom and kept going.
Ten years on, Recursive Thinking has become coaching in every direction people have asked for: code, craft, portfolios, career navigation, and the occasional hard truth delivered kindly. More than 130 designers and technologists have come through it, and it is the longest running thing I have ever led. Every one of those conversations has made me more certain of the same thing: everyone deserves a mentor, no one should have to navigate a career alone, and everyone deserves a straight answer about how the corporate world actually works.
Everything on this page is one idea wearing different clothes. Point at a destination worth reaching, earn enough trust to hear the truth, build the systems that carry the judgment, and give people what they need to go further than anyone would have assigned them. It works on a design system, it works on a team, and it worked on fifty people in a living room. I am excited to find out what it works on next.