Nexus

Nexus is an early-stage startup in Austin that builds connection intelligence for organizations. It uses AI to suggest the right introductions between people inside a membership or community.

I help out as a weekly reviewer. My job is to review the design and user experience across two areas: the public marketing website and the product dashboard. I do not own the roadmap or the code. I give recommendations, and the team decides what to do with them.

The first issue was consistency. The marketing website and the app had grown apart, with different buttons, colors, and layouts. When a user moves from the website to the app, it feels like two different products, which lowers their trust right when they are deciding whether to commit. I recommended a shared set of components and design settings so both areas use the same building blocks.

The second issue started from data. The sign-up process had many people starting and few finishing, and it was one long form with no sign of progress. A long form with no progress bar gives people no reason to believe they are close to done, so they quit. I suggested a progress bar and a profile card that fills in as the user types their name, so their input produces something they can see. The team built it, and the sign-up now shows progress at each step instead of hiding how much is left.

A common theme in my reviews is places where AI was used to generate a screen instead of design one. My view is that AI should make the work faster without lowering the quality of human-centered design. When a screen looks generated instead of designed, the cost does not fall on the team that shipped it. It falls on the user who has to figure it out.

The hard part of this role is not finding the problem. It is getting a recommendation accepted when the only thing behind it is the reasoning. I lead with the user's experience and the cost of leaving it broken, because that is the argument that works on a team that does not report to me.