Fast-tracking client enrollment.

Compass supports families with low incomes to build assets and financial capabilities as a pathway out of poverty. The Family Self-Sufficiency (FSS) Program is voluntary for families in federally subsidized housing.

The internal Compass Enrollment Portal — mobile-friendly, web-based, on-demand — already lets prospective clients learn about FSS, sign enrollment documents, and schedule their first appointment with a financial coach. Its success spurred development of an external version as an add-on tier on FSS Link.

The project's purpose: gather qualitative feedback from internal portal users and prospective external users to understand enrollment needs, determine the most crucial features, and guide design of the FSS Link Enrollment Portal — while informing internal improvements too.

I joined the sprint mid-flight, replacing a teammate who had already worked with the client. Requirements and discussion guides were in place; recruitment had started. I caught up by absorbing project documents, prior deliverables, and internal context — then took the research forward.

Synthesis fed persona work and a feature-prioritization view of what the external portal had to do on day one vs later. Low-fidelity designs went to the hi-fi team to carry forward.

Picking up someone else's sprint mid-flight is a different skill than starting clean. The risk isn't doing bad research — it's doing research that contradicts decisions the client has already made for good reasons you don't yet know about.

I spent the first three days reading everything: prior interview notes, the original discussion guide, the client's internal Slack threads I was added to, the existing portal's analytics, the recruitment screener. Only then did I start adjusting the discussion guide — and only where the new questions earned their place by retiring a specific scope risk.

From there I ran qualitative 1:1 interviews with both internal-portal users and prospective external partners. Two audiences, one synthesis frame, because the whole point was to surface where the experience could be one product with two front doors versus two genuinely different products.

Two findings reshaped the build:

I handed the hi-fi design team a feature-prioritization matrix (must-have on launch vs phase two), persona definitions tied directly to interview quotes, low-fidelity wireframes for the branching enrollment flow, and a measurement plan for the first 90 days post-launch using HotJar and GA so the team would know quickly whether the redesign was actually working.