Banking Centre Monitoring Dashboard

Role
UI/UX Design Lead
 

Overview

I led the UI/UX design to define a banking centre monitoring dashboard using metrics and KPIs previously gathered and prioritized by stakeholders and validated by engineering.

This deliverable was intended to reflect the ideal state of a dashboard and included metrics that were both in and out of scope for a first MVP release. Throughout the process, I collaborated closely with the PO, two PMs, and two engineers to align with business needs and ensure product feasibility. We also collaborated with leads from a third-party vendor to better understand platform capabilities.

Note: To maintain client confidentiality, I have white labelled the product and removed any reference to the company name.

Challenge

In the organization, no dashboard currently exists that provides the health for the entire network, region, province or individual Banking Centres. Current health monitoring is reactive, not proactive and information of the Health of Banking Centres is not widely accessible or available across Technology and Business.

Health of the Banking Centres is determined by Sev notifications, escalations and TSC tickets which offer disparate and limited metrics on hardware, applications, network, issues/incidents, facilities, and more.

Project Goals

• •     Dashboards are available 24/7, identifying the status of the entire network, region, province or individual banking centers.

• •     Health of the Banking Centres is determined by metrics and not dependent on user escalations and calls.

• •     Health metrics are holistic and consolidated addressing all technical issues including hardware, applications, network, issues/incidents, facilities, etc.

• •     Health metrics address high-level watermarks that allow time to address future issues (i.e. network bandwidth/capacity, desktop and laptop speed/capacity, etc.)

• •     Predictability and forecasting of incidents so that teams can action on preventative measures before incidents occur.

Solutions

Understanding metrics
There were several metrics gathered from stakeholder interviews held prior to me joining the team and documented into excel sheets. I used this as a basis for creating the visuals, and worked with the team to improve my understanding of the data and how I might categorize them into high level areas and visualize the data in a way that is meaningful and useful to the target audience. The primary categories that I organized the data into are Location, Applications, Incidents and Insights.

It also took me a few tries to find a solution I was happy with when trying to visualize some of the more complex sets of data. This got me to take a closer look into existing dashboards and design systems such as the IBM Carbon Design System that had several great reference points for functional, accessible and intuitive data visualizations.

Broad target audience
The target users for this dashboard varies from high-level executives to engineers. This meant that the dashboard has to display the data in a way that is meaningful and understandable to users with both business and technical backgrounds. As the user navigates through the dashboard, I wanted to create a layout for each category that begins with high-level information which becomes more detailed as the user moves further into the page.

Accessible data visualizations
Accessibility was a strong priority to myself and the team and I wanted to ensure that the data displayed could be accessible to all users. This included using colours that meet contrast AA accessibility requirements so that text remains legible and color-blind simulation tools to preview how neighbouring colours look to individuals with color blindness.

I also implemented accessibility best practices specific to data visualization such as using different shapes for the data points along a line graph for easier differentiation for situations with low colour contrast, and confirming that the third-party platform is able to display the data in table format for tabular browsing with screen readers.

Insights and forecasting
Predictive analytics is another category that I explored using elements from dashboarding tools such as Power BI and Splunk. Although out of scope for an initial MVP release, the team wanted to explore what capabilities were available and how these could be incorporated into a dashboard view.

This included AI capabilities that could analyze data and make predictions about future trends in number of reported incidents, key influencers that cause incidents to increase, and a Q&A tool that could parse specific queries and display relevant datasets. These predictions can be used to inform decision-making and to mitigate the occurrence of incidents.

Reflection

This project was quite challenging to navigate in the beginning as I was unfamiliar with the data and terminology used, as well as designing data visualizations that are not only accessible, but understandable and provide value to a broad range of user profiles.

It was incredibly helpful to have a supportive team and engineers involved throughout the process to provide feedback as I designed the screens. It was also helpful to be able to establish an ongoing collaborative relationship with third-party platform vendors that were open to setting up calls and walking through technical capabilities in detail.

Outcome

As of now, the team is currently exploring out-of-box and custom solutions from multiple platform vendors. The first MVP of this dashboard is scheduled to launch in Q1 of 2023 followed by several phases to incorporate data from all internal monitoring applications.