As AI technology continues to advance in Silicon Valley, there is a growing need for AI/ML experts. Unfortunately, the supply of these professionals does not meet the demand. Our team recognized this issue and set out to create a recruiting tool that simplifies the process of sourcing AI/ML experts from overseas. This tool aims to match these experts with American companies that require their skills.
It all starts with a simple search on LinkedIn. However, to make it easier for recruiters to save a prospect's information, we have developed a LinkedIn Chrome Extension that automatically scrapes candidate data and shares it across the sourcing agency to create a comprehensive database. Even if the candidate you found isn't suitable for the current position, they may be perfect for another role your colleague is trying to fill.
The entire team will collaborate to source effectively
Once a job posting has been created, utilize Machine Learning to analyze the job description and generate a boolean search for recruiters. This can be easily pasted into the recruiting tool, allowing them to see the matching score instantaneously. The match will apply to the position they are sourcing for and any other job the organization may be seeking to fill.
How easy and intuitive is it to search through the talent pool? (On a scale from 1 to 10, how many people rate over 7?)
76%
How easy and intuitive is it to edit talents’ data?
(On a scale from 1 to 10, how many people rate over 7?)
89%
As a newcomer to the industry, I lacked sufficient domain knowledge to make informed design decisions. To overcome this, I interviewed recruiters and sourcing agency decision-makers to understand the industry better and create personas for our tool.
During my research, I discovered the potential of data visualization and decided to create a talent map using our database with D3.js. I wondered whether a visual representation of talent locations would provide more information for our organization to plan its hiring strategy. Although our investor was intrigued by the concept, we were unable to move forward with this feature due to the limitations of our data as a startup.
Although machines are powerful, they are not always accurate. To overcome this challenge, I created a tool to help recruiters manually correct the labeling for resumes and feed it back into the machine to improve its learning capabilities. This feature only has long-term benefits, but we didn't pursue it at the time.
We aimed to work in a Lean style. As I am accountable for both UX and Front-end, I streamlined my workflow by creating a prototype template using HTML+ CSS. I then conducted usability testing on it, so that I can implement the changes faster. This is possible since I need to work directly with the back-end once I have finalized the design.
Lack of Information
No motivation for recruiters to maintain
Searching experience is too complex
