Led Design for Predictive Risk Platform, Securing Major Clients
Gideon Awolesi
The Merkle Science platform predicts cryptocurrency risks and gathers intelligence to assist crypto companies, banks, and the government in finding, investigating, and stopping illegal activities involving cryptocurrencies. The company provides services for monitoring transactions, investigating crimes, conducting background checks, and teaching employees about compliance.
My Role:
As a designer, I was in charge of designing and building user-centered solutions. This required working closely with all parties involved to fully grasp and resolve tricky issues related to following crypto rules and stopping crime. We worked in a sprint model and used Jira to track requests and requirements.
Process and Approach
1. User Research: Understanding Client Needs
To deeply understand the needs and pain points of our clients, as soon as I joined, I initiated an extensive user research project. The goal was to gather qualitative and quantitative insights to inform design decisions. I wanted to know what the current status was, as there was not much design work being done before. I needed to know who our user was and the personas defined as soon as possible.
The first step was an initial workshop that involved key stakeholders, including product managers, developers, and marketing teams. The objective was to outline the user research's primary goals and identify key focus areas. The workshop highlighted critical challenges such as data overload for compliance officers and the need for detailed transaction histories for forensic analysts.
Everyone's input shed its own light to paint the larger picture I needed to understand what to fix.
The next thing I did was deploy a survey to target the client base. This was the quantitative part of the research, and more than 100 responses were collected. The survey was designed to gather quantitative data on their workflow, challenges, and satisfaction with existing tools.
The final step was in-depth interviews to build on the survey results. I conducted 20 in-depth interviews with 10 compliance officers and 10 forensic analysts. Each interview lasted about an hour, providing deep insights into their daily routines, pain points, and specific needs.
Problems Identified and Solutions:
Compliance Officers:
Forensic Analysts:
The results were that these solutions cut the time that compliance officers spent managing data by 30% and made the work of forensic analysts 25% more efficient. Surveys sent to the users two weeks after the solution's implementation served as the basis for measuring this outcome.
2. Competitive Analysis: Standing Out in the Market
Next was a full analysis of top platforms like Chainalysis and Elliptic to ensure Merkle Science stood out from the competition.
I conducted a SWOT analysis (Strengths, Weaknesses, Opportunities, and Threats) for each competitor. This included examining features, ease of use, and customer feedback.
This led me to a lot of findings, but I will summarize them into two main points:
Chainalysis has powerful forensic features but a complicated and hard-to-use interface.
Elliptic: Easy to use, but not many ways to customize.
Based on this, I listed a couple of solutions designed to achieve the following outcome:
focused on making an interface that is easy to use and gives clear, actionable insights.
Put in place behavior-based transaction monitoring to find illegal activities before they happen, which sets us apart from competitors.
After implementing them, we got more recommendations by our existing user base to other clients, and we started running trials with Binance, Crypto.com, and even a live testing session with some agents in the US GOV.
3. Wireframing and Prototyping: Iterative Design for Optimal Usability
After all the data collection, it was time to start drawing pixels. I set up designs and changed them to ensure they worked and met users' needs.
Process:
Wireframes: The first were made with Figma, focusing on layout and flow.
Prototyping: I used Figma to make interactive prototypes to test their usability.
Usability Testing: Ran 5 rounds of usability tests with a group of 15 people, some of whom were forensic analysts and some of whom were compliance officers.
Findings:
The first designs had trouble with navigation and call-to-actions that were not clear.
Users liked data that was shown in a more visual way.
Actions Taken:
Made it easier to navigate and improved the visual hierarchy.
Added visual data displays like charts and graphs to help people understand the data better.
Outcome: Usability improvements led to a 20% increase in user engagement and a 15% reduction in task completion times.
4. Design System Development: Ensuring Consistency and Scalability
After completing the immediate fix needed, more design requests were coming, and it was important to
Process:
Design System: Set up a flexible design system using Atomic design principles.
Collaboration: Closely worked with front-end developers to ensure smooth integration.
Challenges and Solutions:
Problems with Compatibility: At first, the design system and frontend technology did not work well together.
Solution: Made different parts of working with the front end's limitations and kept in touch with developers to make changes in real time.
The flexible design system made it possible to implement consistent and scalable designs. This reduced design update time by 40%, which was a big deal.
Outcomes and Impact
High-Profile Client Acquisition:
Got well-known clients like Binance and Crypto.com.
Regulatory bodies in the U.S. were interested, and live tests with the FBI were conducted.
Scalability:
The platform handled millions of requests, demonstrating its dependability and expansion capacity.
73% increase in usage and 90% increase in use of previously unexplored features on the platform.
Clients liked how easy the platform was to use and how many features it had. A compliance officer from Crypto.com talked about how the platform made their AML processes a lot easier.
Like this project
Posted Aug 7, 2024
Merkle Science Risk Analytics
Likes
0
Views
3
