AI dashboard
Predictive rate adjustments
the ask
My team was tasked with creating a tool that harnesses machine learning to help facilities predict and implement rates that maximize the money they spend on contract healthcare workers. The process for making these rate changes was originally manual and prone to mistakes. The facility would email their account manager a table with new rates and we would need to implement them each manually in the backend. The vision was to create a self-service tool that allows the user to adjust their own shift rates while also making smart suggestions on how they could optimize the money they were already spending.
workshop
To understand the (very complicated) current process for updating rates, I ran a workshop with one of the backend engineers and created a diagram of how the rates layer on each other.
user interviews
Because of the complexity of this project, I wanted to gather as much information up front as possible. I was able to recruit 3 users to interview.
current state user flows
To help me better understand the current state and validate my assumptions, I created a diagram of the current user flow based on the information I gathered from user interviews.
new user flows
From there I created new user flows and documented potential entry points to pitch to my engineers.
usability testing
After I felt I had a good solution, I wireframed and prototyped my idea. I initially sent my usability test out to internal teammates to gather their feedback before it went out to users. I gathered feedback from different teams including client facing teams and engineering teams and was able to implement the insights before it went out to users via our user testing platform.
the results provided valuable insight into areas that needed more work
handoff
After synthesizing and documenting the findings from usability testing, I created high fidelity designs to hand off to developers. I collaborated with them and supported them throughout the entire development process.
results
This feature saved hours of time for our internal team. We also had an overwhelming response from users who were excited to have more transparency around how rates are calculated and how redistributing their funds can optimize their money.