Project Haystack, 2020
Brief:
Bring to life our vision of an investigative dashboard tool that aggregates, surfaces and visualises many data sources to enable investigators to identify patterns of fraudulent activity.
Lead Product Designer, 100% remote
Role:
End to end UX/UI, data visualisation, pitch materials
Activities:
Background
A wide range of industries use investigative technology to identify fraudulent and criminal activity. The finance sector uses solutions to combat money laundering and the government uses them to monitor the illegal import of good across borders.
The problem
There are a range of off-the-shelf solutions that can help agencies conduct their investigations, however they have some aspects that could be improved to enhance the end user experience. Current products:
-
tend to be a one-size-fits-all solution
-
use multiple interfaces to surface data
-
are limited in data sources
The solution
The data and insights team at Capgemini identified a gap in the market for a dashboard solution that enables fraudulent behaviour to be investigated seamlessly with a product that pairs federated data, with a customisable tech stack and a slick front-end view. The team enlisted me to help bring their vision to life by creating a value proposition complete with full UI design and pitch deck to gain investment for the final solution to be built.
Understand and define
To truly understand the problem, I conducted one-to-one (video) interviews and observational research (screen share) with various SMEs to understand the current products on the market and the scenarios that come up when using them.
I identified 3 key user groups: the investigator, the analyst and the data manager, and mapped out their journeys to ensure I was thinking about them at all times when later designing the product flow and UI.
I also spend some time exploring existing tools such as Linkurious to become familiar with the UI, and also spent some time with data scientists to understand the back-end set up. This helped me to gain insight into the different data sources and which elements needed to be surfaced and visualised at what point in time.
Data visualisation
The key data sources I had to visualise within the solution are:
-
ANPR data
-
Phone records
-
Financial data
A secondary group of data sources included additional intelligence such as references to places or peoples that might be found in BBC news articles for example, and this was correlated with other data through natural language processing.
WIP high level journey
Scenario mapping
I used one of the common scenarios that came out of the initial discovery to form the basis of the user journey through the UI and the narrative from the product demo - an instance of illegal import of goods by ferry from Ireland.
UI
I started off by sketching out low-fidelity wireframes to gain feedback and iterate quickly, before turning the designs into high-fidelity in Figma.
.png)
Communicating the proposition
To convey the product effectively to key business stakeholders, I created an animation (below) detailing how a user would navigate through the interface to investigate a case. The animation also details how data would be aggregated and surfaced.
What next
The team gained internal investment to build out an MVP of the solution, however it did not go any further than this due to budget restraints.
Read another case study
