Merble Redesign
 2026    Group project

Merble is an AI-powered lending intelligence platform for community banks and credit unions. It detects real-life signals: job changes, property value shifts, credit improvements, and helps loan officers reach the right member at the right moment, before they seek financing elsewhere.


Category: Product Design, UI/UX Design
Role: Product Designer
Responsibility: Merble - Analytics Overview & Performance
Skills: User Research, UX Strategy, Information Architecture, Interaction Design, Data Visualization, Wireframing & High-Fidelity Prototyping
Tools:  Figma,  Miro,  Adobe Illustrator, After Effects, Claude, Figma Make



    Our team was responsible for designing the Analytics & Accountability section, the part of the platform that answers one question for a credit union manager: did Merble pay for itself this quarter?

    Within the team, I was responsible for the Analytics Overview and Performance pages. The Overview serves as the daily dashboard, surfacing the most important signals, funnel health, and LO performance at a glance. The Performance page gives managers and supervisors the tools to evaluate loan officer accountability, campaign ROI, and structured alerts structured by severity and type.



    Problem

    The existing analytics page had real data but communicated none of it effectively.



    Pain Points


    No clear hierarchy. There was no visual priority telling the manager where to look first. The most important information was buried at the same level as everything else, and the dark color system made the page even harder to read and navigate.

    No Actionable Insights. The information stopped at showing the raw data, there was nothing to help users understand what the numbers meant, no comparison to the market, and no direction on what to do next.

    Data Without Visual Contexts. The page was dense with figures and texts but lacked any visual representation like graphs or charts, making it harder for users to quickly identify patterns and draw conclusions at a glance.



    How Might We
    How might we design an analytics experience that helps credit union managers instantly understand whether Merble is generating value without having to interpret raw data themselves?

      
     Information Architecture

    Rather than consolidating everything into a single analytics page, we structured the section into four distinct sub-pages.

    The Overview page acts as the daily dashboard, surfacing the most critical information a manager reaches for first: total signals, campaigns sent, total engagement, and loans closed, followed by the conversion funnel and a summary view of signal and LO performance. It gives the full picture in one scroll without requiring any navigation.

    The Performance page goes deeper into accountability, a detailed breakdown of how individual loan officers and campaigns are performing. It includes a performance alerts section powered by AI analysis, which proactively flags what needs attention, categorized by type and structured by severity so managers always know where to focus first.

    The Signal Deep Dive and AI Intelligence pages give marketing leads and analysts the tools to investigate individual signal performance and access AI-generated recommendations and weekly summaries.

    This structure ensures that a manager who only has 30 seconds gets what they need from the Overview, while a supervisor who wants to dig into LO accountability or campaign ROI has a dedicated space to do that without cluttering the top-level view.



    Rapid AI Prototype




    SolutionDesign System







    Key Features



    Headline numbers at the top of the page giving managers an instant snapshot of the full pipeline before they scroll into any detail, with the most important metric highlighted to immediately communicate what matters most.


    A step by step visualization of the member journey from Signal Detected through to Closed, with each stage plotted against an industry benchmark so managers can see not just where drop off happened, but whether it's actually a problem compared to the market. At the top of the funnel, the biggest drop off stage is automatically surfaced with an AI generated recommendation telling the manager exactly which step needs attention first and what to do about it, before they even read the data below.


    Performance alerts structured by severity and filterable by type, the alerts section surfaces what needs attention before the manager has to go looking for it, with each alert paired with a specific recommended action so there is always a clear next step.



    Final Design


    Analytics Overview


    Analytics Performance


    Reflection This project taught me a lot about making design decisions that serve a real business purpose. Through regular meetings with the founder, learning about the product, and researching a field I was completely unfamiliar with at the beginning, I gradually became more confident in understanding both the industry and the users we were designing for. By the end of the project, I felt genuinely invested in the problem we were solving.

    One of the most valuable parts of the process was designing on top of an existing product rather than starting from scratch. Instead of immediately focusing on visuals, we first had to analyze the weaknesses of the original experience, identify user pain points, and understand why the information was difficult to navigate. This made the design process feel much more intentional and strategic.

    Collaborating with the team was also one of the most rewarding aspects of the project. Regular discussions, feedback sessions, and aligning on a shared design direction made the work feel much more team driven and strengthened the final outcome.

    This project also introduced our team to using AI as part of the prototyping process. AI helped us quickly visualize ideas and communicate concepts early on, but it also showed me the importance of not relying on AI generated layouts too heavily. AI often prioritizes quantity of information over clarity, which taught me to think more critically about hierarchy, prioritization, and intentional UX decisions.

    The biggest takeaway from this project was realizing that good analytics design is not about showing as much data as possible, but about understanding what information is important and presenting it through a clear visual hierarchy.






    © 2026 YANING HU - All Rights Reserved.