Noble's Policy Builder and Log
Turning Black-Box Decisions into Actionable Insight
MY ROLE
UX flows & user scenarios
UI design
Prototyping
Design system contributions
Noble’s Policy Builder enables businesses to generate automated credit decisions—but users lacked visibility into why those decisions occurred.
Noble simplifies how companies make financial decisions at scale.
It automates onboarding, underwriting, and monitoring in one unified platform.
At Noble, policy outcomes were technically correct—but operationally opaque. Underwriters could see decisions, but not understand or act on them. This created a hidden dependency on engineering and slowed critical business workflows.
I led a redesign of the Policy Results experience to transform it from a static output into a decision intelligence layer—making system logic legible, actionable, and scalable across use cases.
Key Impact
I designed a dynamic results interface that exposes both decision outcomes and the logic behind them.
-
Increased transparency into automated decisions (Approved, Pending Review, Rejected, Error) with data inputs tied directly to outcomes and pass/fail visibility
-
Enabled self-serve troubleshooting by surfacing root causes of failures, reducing cognitive load and time spent interpreting the results
-
Consolidated complex data into a single, digestible interface for auditing and edge-case analysis while also providing access to underlying logic (e.g., JSON) for advanced users
Key Design Move
A drawer (side panel) system allows users to explore detailed data without losing context—keeping the main results view intact while enabling deep inspection.
Core Problem
Lack of Visibility Into Key Credit Decisions
The original experience surfaced results—but not reasoning.
Users couldn’t:
-
Understand why a policy failed or stalled
-
Investigate “Pending Review” decisions
-
Diagnose errors or rerun policies independently
This resulted in high cognitive overhead for underwriters who needed to make high-stakes credit decisions quickly and confidently.
Product Strategy
Quick action, close collaboration, & in-flight testing
Rather than simply “improving visibility,” I reframed the experience around a higher-order goal:
Make system decisions explainable, traceable, and actionable at every level of abstraction.
This led to three guiding principles:
-
Legibility over simplicity
Complex systems shouldn’t be oversimplified—they should be interpretable. -
Progressive disclosure as infrastructure
Information should scale from high-level outcomes → granular logic without breaking context. -
Actionability as the success metric
Insight is only valuable if users can do something with it immediately.
The solution was validated through strong internal feedback and positive user response, demonstrating clear improvements in usability and clarity. It successfully translates complex system behavior into an intuitive interface, while also establishing a scalable foundation for future self-serve capabilities.
Summary
Evaluate, Troubleshoot, Act
Rather than forcing users to interpret static outputs, the redesigned experience enables users to move beyond simply viewing outcomes to actively understanding and acting on them. By making decision logic transparent and navigable, it supports faster, more confident underwriting without relying on engineering or support.
What Users Can Now Do
-
Trace every decision back to its originating data
-
Identify exactly where a policy failed
-
Resolve certain issues immediately
-
Rerun policies with confidence
Note: Feature impact metrics were not captured prior to my departure.
Challenges & Tradeoffs
-
High system complexity required deep collaboration with engineering
-
Early concepts (e.g., progress bars) proved unscalable and space-inefficient
-
Tight timelines led to premature reliance on UI components over foundational exploration