Salary Intelligence for everyone

As a senior designer in an HR product team I was tasked with developing the user experience for a new technology that used AI to create salary recommendations with an accuracy previously only possible with expensive subscriptions to providers who hold massive but not frequently updated data sets. Getting accurate recommendations required some complex interactions and our system had to accommodate both people who already have expertise in salary analysis, and current users of our HR system who might not have any previous experience in this area. We used formative and summative qualitative research, rapid prototype generation, and close collaboration in a cross functional team to develop a working alpha product that even low confidence users could use to generate accurate and usable salary recommendations.

I interviewed both SMEs within our company and customers representing the various levels of confidence we needed to accommodate, to identify if there was a market opportunity for this new capability, and to identify the possible strategies and behaviours we would need to design for. We used tried and true methods like lean business canvases, and qualitative semi structured interviews to test the assumptions we generated in our early explorations.

Once we had established that there was a good case for prototyping the recommendation engine as a product, I collaborated with our engineers and data experts to break down the data we could generate into consumer friendly and market appropriate insights. I also provided initial high level UI guidance and mapped out the main flows and UX considerations to guide the engineering team as they developed a technical proof of concept.

Using our existing design system as much as possible I developed Figma designs for the two main interactions that form this feature; selecting the position to be analysed, and manipulating the data in the resulting analysis. I used dummy data to build an AI generated prototype in Figma Make which was used to test the complex interaction involved in identifying the position to be analysed, and hand made design prototypes in Figma to test basic usability and comprehension. We collaborated with the engineering team to ensure feasibility of the designs as several new design patterns were required outside of our design system, and the prototypes were tested with users and reviewed for compliance to the design system and for allowing people to create salary recommendations.

The recommendation engine has just finished testing and is scheduled to be integrated into the main product with our new user interface over the remainder of 2025, with a beta testing version available to select customers in 2026.