Recruiting with AI

By Nikki Heaston

This week our class was introduced to Frrole DeeperSense. Frrole specializes in building the social web that can provide clients with predictive consumer insight. DeeperSense is a product of Frrole that offers “Intelligent Hiring.” The sub-heading on DeeperSense’s main webpage reads “Hire smarter and faster using AI-built candidate profiles.” The process is simple: enter a job candidate’s Linkedin URL or social media handle and a summary of that candidate’s personality characteristics will be written. This can give insight into who this person really is outside of the carefully-chosen, search-optimized vocabulary, and dazzling accomplishments listed on his/her resume. This technology is a lifesaver to hiring managers, especially in large firms, who receive hundreds of applications per job. The sheer volume of resumes and social media accounts to sift through would require a great deal of patience and overtime pay. Now with AI like DeeperSense recruiters can quickly and easily review candidates and make more informed decisions about finding the right fit for their teams. If you need a more specific recruiting tool, truck drivers, for example, you will need to go through a different agency. Recruiters may want to see here about this different option so they can get started on recruiting the people that they need.

But how accurate is this AI really? Can it give your dream company a realistic picture of you? And if you’ve only been an active user on a particular social media platform for a few months, is your data skewed compared to someone who has been using that platform for several years? Wouldn’t that also give you an unfair advantage? Perhaps using international employment specialists would be more effective?

To test out DeeperSense’s accuracy I entered my LinkedIn URL and my results were this:

My initial response was “Wow, this is super accurate!” Then I decided to run a little experiment and search my husband’s LinkedIn URL. And here were his results:

Tim and I are very different people with very different backgrounds. I majored in Business in my undergrad, he majored in English/Creative Writing and German. My work experience is in retail operations, his is in advertising and marketing. My job descriptors say: Retail Store Manager, and Pricebook Analyst, Tim’s job descriptors say: Copywriter, Editor, Video Producer, Workflow Manager, and Freelance Writer. However, on DeeperSense it appears that he and I are almost the same job candidate. Broken down, the profile consists of 11 different descriptors. Here is a side by side comparison of our profiles:

Of those eleven, Tim and I have six that are verbatim the same (highlighted in yellow). Three are completely different (highlighted in green), and two of them are half true for both of us. However, when you look further you see that even some of the differences are only slight:

• Tim can be critical but is always thorough vs. me who likes to take a systematic approach and values quality and accuracy.
• Tim can be nervous at times vs. me who is a little anxious sometimes.
• Tim has a sharp goal focus and doesn’t miss them when he takes them on and I achieve results and complete tasks that I take up.

I’ll admit that during class I was concerned about the subject of using AI software to screen job candidates. I didn’t believe that hiring managers would get an accurate picture of me – that AI would state things about me that were really correct. Now, I’m still concerned, but for a different reason: that AI software is only helpful in giving broad generalizations about candidates, making us appear to be basically the same. Whilst this is true, the software is mainly used to try and whittle down the applicants so that the hiring managers can find the best candidate. There are so many pieces of software around these days to help make business processes easier (like screening agencies such as Checkr who have platforms to make background checks efficient), so this is why businesses are starting to make use of them. Software developers are constantly trying to advance their software too, making sure that they have more of an advantage over other competitors. Some software uses Mirantis for container orchestration. This allows developers to scale and run the software for clients.

Obviously, managers won’t look at this alone in their hiring decisions (yet), but if the decision to hire me or Tim to be on a marketing team, for example, how much insight does this really give? If you’re paying for this product, are you seeing your ROI?

And let’s not leave out the issue that Amazon faced when it ran its own secret AI to recruit top talent. This is a headline from Business News in Reuters: “Amazon scraps secret AI recruiting tool that showed bias against women.”

According to the article: “Everyone wanted this holy grail,” … “They literally wanted it to be an engine where I’m going to give you 100 resumes, it will spit out the top five, and we’ll hire those.”

“But by 2015, the company realized its new system was not rating candidates for software developer jobs and other technical posts in a gender-neutral way.”

“That is because Amazon’s computer models were trained to vet applicants by observing patterns in resumes submitted to the company over a 10-year period. Most came from men, a reflection of male dominance across the tech industry.”

“In effect, Amazon’s system taught itself that male candidates were preferable. It penalized resumes that included the word “women’s,” as in “women’s chess club captain.” And it downgraded graduates of two all-women’s colleges … ” Additionally, “Amazon declined to comment on the technology’s challenges, but said the tool “was never used by Amazon recruiters to evaluate candidates.” The company did not elaborate further. It did not dispute that recruiters looked at the recommendations generated by the recruiting engine.”

So what does this mean for you and me, job seekers with no power over the processes used by recruiters? What are we to do with imperfect AI that some companies are forging ahead on using? Forbes offers four pieces of advice here, but the most important is to remember to be honest and authentically you.

Twitter: @HeastonNikki, Instagram: nikkiheaston, LinkedIn: https://www.linkedin.com/in/nikkiheaston/

Featured image:
Photo by Franki Chamaki on Unsplash