Context
WSIB (Workplace Safety and Insurance Board) provides wage-loss benefits, medical coverage and support to help people get back to work after a work-related injury or illness.
Our team was tasked with designing a hi-fi prototype of a GenAI-powered search engine that minimizes the difficulties and search time needed for Return-to-Work (RTW) specialists to look up accommodations for injured workers.
For this project, my impacts include (but are not limited to)
ideating and defining the key frames and main flow
designing and reiterating the accommodation editing screens
collaborating with developers to ensure quality POC delivery
Due to NDA, some details are generalized. Reach out to learn more!
CoMPANY
WSIB
role
UX/UI Designer
timeline
Sept - Dec 2024
team
1 Product manager
2 UX/UI designers
2 Developers
1 Comms specialist
The problem space
Return to Work (RTW) Specialists support injured workers throughout their recovery and get them back to work in the fastest, safest ways possible. A key part of their process involves recommending accommodations to integrate into the workers’ day-to-day processes. However, there is a big challenge.
Fragmented and Time-Consuming Search Process
The current process to search for these accommodations is cumbersome, requiring manual filtering through an excel sheet. Additionally, these accommodations are stored in multiple forms of media (images, text, PDF), making it difficult to quickly populate a structured database.
Understand the user
This project initially started as a one-week sprint challenge. Due to the tight timeline, we were unable to conduct formal user interviews. Instead, we asked our design leads to roleplay as an RTW specialist and help us better understand user needs. Here are our findings:
user's goals
Efficiency
Find and recommend appropriate work accommodations efficiently
ensure safe return
Ensure injured workers return to work safely and quickly
maintain records
Maintain accurate records of accommodations for future cases
pain points
Long search process
Searching for accommodations requires manually filtering through an Excel sheet, leading to inefficiencies.
scattered info
Accommodations exist in many media formats and may be outdated, making it hard to find every information in one go.
time constraint
Managing multiple cases means limited time to research accommodations for each worker
When this turned into a full-term project, our PM and Comms team spoke more to stakeholders for more information on features we may have missed
Starting the design
The design team started by writing down all the key frames we needed to figure out the user flow. We took a few minutes to make a quick lo-fi version of each frame to ideate and see what the important elements are.
That helped guide us to make our first hi-fi prototype.
As we discussed the features of the web app and looked back to user needs, our PM proposed creating a mobile version, so RTW specialists can support injured workers anytime, anywhere. Given the smaller viewport, we carefully defined which features were best suited for mobile versus web.
I explained that editing accommodations on the mobile version would be impractical, as it requires viewing accommodation documents and navigating multiple form fields. As a result, the mobile experience was focused on quick and efficient accommodation searches.
Refining
We went through a total of 8 versions, reiterating and designing additional screens as we continued defining the user flow and gaining feedback from stakeholders and design leads.
We designed new global components and enhanced the existing design library to improve consistency and accessibility across the platform.
Solution
Easily search for accommodations
A streamlined, GenAI-powered search experience that allows specialists to quickly find relevant accommodations using key words, tags, and filters.
Upload & Auto-Summarize Documents to update database
New accommodation files can be uploaded to quickly find all accommodations and provide a summary of it.new accommodation files to be automatically summarized by GenAI and structured into the database.
Build trust with ai content through human verification
Before being released internally, all AI-generated accommodation summaries enter a verification backlog, where RTW specialists review and approve them for accuracy.
Takeaways
Lessons
Cross-functional collaboration
Throughout the project, I collaborated closely with developers to design components that they could be build. This constraint helped improve my communication skills and pushed me to advocate for users through each iteration.
show relevant info where needed
Our design leads provided feedback throughout the project. One key advice was to surface important information when relevant. The “Contains AI-generated content” label should only be shown when AI content was displayed to reduce user confusion.
Next steps
This project was presented to stakeholders and got approved for a full-scale investment! It will be handed-off to the next cohort of interns.
As most of our feedback came from stakeholders after seeing a run-through, usability studies must be conducted to better optimize the user experience.
Revisit the roadmap, as there were some features we could not include due to time constraints, but recorded them for future versions
Update: this project has been shipped internally!