Designed the Adaptive Network Fulfillment product that allows retailers to compose, orchestrate and monitor their fulfillment strategies. Its powerful data visualization capabilities and simple rules-UI aims to maximize the profitability using an underlying sophisticated optimization algorithm.
High-Level Product Goals
- Enable means to expand the fulfillment network beyond traditional Distribution Center and Warehouses, to thousands of stores, while keeping in mind the special considerations unique to store
- Device capabilities to enable retailers to meet the new expectations of customer, get the products fast and free, while still increasing the profits
- Build mechanisms to prepare for spikes during special events like holidays, flash sales and promotions
- Enable usage of entire network inventory to optimally source
Design Level Goals
- Interfaces that are easy to configure, use, change and understand
- Solution that is scalable, flexible and hides the complexity of the underlying system design and algorithm
- Easy to troubleshoot and monitor the system performance. UIs to help build trust in the new means of optimizing fulfillments
- Visualization of the results and reports highlighting key metrics for network performance e.g. project the revenue, the savings and the customer service
Initial Challenges
- The project was short of 60 development days on UX involvement Day-1.
- Being the core driver of Omni application, this project was high profile in nature. With the departure of Product Manager early on, and a very tight timeline this project posed a steep climb from the day 1 in the discovery phase
- There was a steep learning curve in understanding the original allocation eco-system.
- Product Management, Architecture, Development and Research teams were working on different phases of the project and they were not aligned.
UX Process
Designing UIs for a highly flexible optimization algorithm, from scratch, required me to take a deep dive into areas which were highly technical and complex. I researched the current state of the algorithm, understood the initial high-level expectation, conducted my own research in understanding the problem, fully internalized an evolving highly sophisticated optimization algorithm of the science team. With limited access to the end user and limited expertise in the domain, there was a lot of self-initiated learning from in-house experts. The evolving algorithm capabilities, the gaps with the high-level goals and what was capable with the technical team required me to adopt the dual-track agile process. I took charge of project planning to ensure we are delivering as we are discovering.
With fastly shaping requirements, it was incumbent on design to help drive the delivery process by fully immersing in the discovery process. Rapid sketching, prototyping and sharing the assets with internal and external stakeholders facilitated collaboration and expedited decision making.
I divided the UX work in the project into 3 groups:
First, screens that were sure bets and could be put into development cadence - This would give a head start on development and save some development days from being wasted.
Second, screens that could get the high-level articulation of project management goals and could help them visualize the end solution and get them to vet it out with the retail customers. For these screens, I worked closely with VP of product management to get quick feedbacks from more end customers.
Third, screens that could help surface the powerful flexibility of the science algorithm to all stakeholder to help in discussions and making adjustments for the end users. For this, I worked closely with Science and Dev team to come up with UIs that help bring clarity and fosters collaboration with other key stakeholders.
Since the target was to provide the screens for development in a continual fashion, I started with the screens that showcase the basic flow of the algorithm.
The project goals shifted significantly through the course and with multiple iterations and feedbacks with the internal and external stakeholders we were able to come up with the critical flow that is required for the customer.
Once the development started working with the critical flows, we started working on the edge case scenarios that needed to be included on the screen.
Here is the glimpse of the Minimum Viable product that the team built in a shorter time duration.
This was an opportunity for me as a product designer, to take multiple additional responsibilities, such as doing gap analysis, understanding the underlying systems analysis and to work tirelessly to bring UIs to the sophisticated algorithm of science team, keeping in mind the domain knowledge expertise of the end user and flexibility expected out of the user. It was a very gratifying experience towards the end.