I was tasked to design Polaris, an incredibly complex financial data system, from discovery to delivery. Polaris is a trade promotion optimization tool that leverages existing data from Compass (the parent trade promotion management tool) to further calculate a key account manager’s promotion data in order to optimize new promotions. It allows key account managers to set their spend and time parameters and uses an IBM machine learning model to enable sales forecasting.
I designed KraftHeinz’s internal applications Polaris (supported by an artificial intelligence IBM API) and Compass (an Enterprise Promotional Data Planning platform). For this case study, I will focus on Polaris.
I designed KraftHeinz’s internal applications Polaris (supported by an artificial intelligence IBM API) and Compass (an Enterprise Promotional Data Planning platform). For this case study, I will focus on Polaris.
When it comes to data visualization, users seem to share a common goal. As the Global VP Design at IBM articualted, "(the users) have data, they have questions, and they need an analytics tools that will help them make sense of their data and turn it into useful business insights, while reducing uncertainty." How might the design reduce redundancy, enables real time interation with data, and assist users in the data analysis process?
After shipping lo & hi wireframes and annotated wireflows, we are currently involved in an iterative process working with the program manager and front end developers to test this product and conduct further user research. We hope to sit with key account managers soon to hold conduct inquiries and learn more about how they interact with the application.