Clinova

About Project

Clinova is an EDC system for running Phase II–III clinical trials: the software clinical data managers and monitors use to capture, verify, and lock trial data across multiple sites. The design problem sits in the tension the domain is built on - the platform has to move fast enough for daily site work while holding a regulated audit posture where every value change is signed, reasoned, and traceable. Getting that balance wrong in either direction breaks the product.

My Role & Activities

• Domain framing & research - pinned down what the product actually was, since "clinical data management" is easy to confuse with electronic health records despite the two having opposite mental models; benchmarked against the category's reference systems and reframed the work around EDC and 21 CFR Part 11 before any layout decisions, a framing that cascaded through every screen that followed
• Information architecture - trial data is deeply nested (study, site, subject, visit, form, field), so the navigation had to make that hierarchy walkable without burying the one thing a monitor is hunting for: where the data isn't clean yet; structured the product around a dual-level model that keeps study-wide context available while letting a user drop straight into a single discrepancy
• Interaction design for regulated states - mapped the full lifecycle of a discrepancy (automatic edit checks, query generation, response, resolution) and designed the form states, audit trail, and electronic signature moments so compliance reads as a first-class part of the interface rather than a layer bolted on top
• Design system & data-density work - built the token architecture with a hard rule separating brand color from clinical status, so the reds and ambers a reviewer relies on never compete with interface accent; dense matrices, longitudinal lab views, and permission grids all draw from the same system

Outcome

Clinova became a tool that lets a monitor open any study and immediately see which subjects need attention, trace a single out-of-range lab result from the chart where it surfaced through the query it raised to the overdue follow-up it triggered, and understand exactly who is permitted to resolve it. The regulated behavior - signatures, reasons, audit history - is woven through the daily workflow instead of sitting apart from it, which is the difference between a system people trust with trial data and one they work around.