C A S E   S T U D I E S

​1. Project Overview

An established loan monitoring product, in the market for over 20 years, recently integrated AI-powered tools for detecting erroneous or potentially fraudulent financial reporting, as well as enhancing financial and margin analysis. To fully leverage these capabilities, the product required a comprehensive redesign to integrate the  AI-driven insights and optimize user workflows for efficiency and clarity.

​

2. The Challenge

The primary objective of this project was to upgrade and optimize three key processing flows for three different types of lender-provided data. Even though, the AI engine significantly improved calculation efficiency and accelerated error detection, we had to deal with varying data formats from different lenders, which introduced complexities in interpreting. Users were responsible for interpreting the AI-generated insights and taking pertinent actions. The need for a redesign came from the users' need for clarity and decision-making efficiency.

​

My role implied identifying and refining requirements, as well as all UX-design related activities throughout the duration of the project.  

​

3. Research & Discovery

To tackle this, I dedicated a fairly amount of time to the discovery phase, gaining a deep understanding of both business objectives and user needs. This involved:

• Industry & Product Immersion: Learning key terminology, business concepts, and the broader financial context.

• Stakeholder & User Interviews: Conducting extensive discussions to identify pain points and workflow inefficiencies.

• Lender Surveys: Gathering insights directly from lenders to understand data formatting challenges and user expectations.

• Comprehensive Competitor Analysis: Analyzing similar products to identify best practices and areas for differentiation.

• Information Architecture & Flow Mapping: Structuring data logically & visualising flows.

​

Documenting all findings and creating user personas laid the foundation for the next steps in the process.

​

Key findings: ​

   a. The array of data formats from lenders created numerous edge cases that needed to be accounted for

   b. Given the large volume of data processed by the system, we had to focus on having an effective way for data visualization

       and building robust, but intuitive error handling features

   c. There were two primary user categories:

• Novice - unfamiliar with the product and its functionalities​

• Expert - highly experienced with the existing version of the product and its workflows

       Personas and user journeys were created for both user categories.

​

4. Strategy

The approach was iterative, working in sprints with biweekly stakeholder meetings that ensured continuous alignment. Each processing flow was tackled individually, allowing us to focus on one aspect at a time. 

I created a concept prototype for user testing, gathering feedback from both stakeholders and users. The feedback was integrated into multiple design iterations, after which we performed usability tests and finally, applying the branding design elements for the vision prototype that went into development.

​

5. Solution & Impact

The proposed solution introduced several key features that addressed core challenges:

• Consistent & Clear Data Visualization: A unified approach to visualizing data across all formats, ensuring clarity regardless of the source.

• Advanced Filtering System: A powerful filtering system that enabled users to quickly isolate and focus on error-prone items, improving workflow efficiency.

• Error Categorization & Visual Indicators: Errors were categorized based on importance, with each error accompanied by unique, easily recognizable visual identifiers and clear instructions for possible actions.

• Helper Mode: A toggle feature that displayed helpful text and tooltips throughout the app, designed to assist novice users without overwhelming expert users, who could disable it as needed.

• Interactive Process Stepper: An interactive stepper that allowed users to navigate through the flow, returning to previous steps while retaining modifications, ensuring flexibility and control.

​

The impact of this solution was reflected in a speedy error indentifications and handling, as well as improved task completion time. The introduction of features like "Helper mode" and the interactive stepper were highly appreciated by users, as it addressed the needs of both novie and expert users, reducing the learning curve, reducing frustration and ultimately making the product more accessible.

​

6. Reflection and Learnings

This project was one of the most challenging I've worked on so far, primarily due to the complexity of the B2B context and the intricacies of the financial domain. Navigating through unfamiliar financial concepts, shaped by legislation from a different country, was definitely a steep learning curve. Additionally, having to balance the potential of AI tools with a 20-year-old implemented logic presented a unique challenge, to say the least.

​

I think that, through this experience, I gained invaluable insights into the importance of flexibility in UX design. I learned that maintaining a human-centric approach is essential, even when integrating AI-tools. Technology serves the user's needs and enhances their experience, so prioritizing usability and empathy for the user remains critical.