Making recruitment fairer and better with AI

AI is going to change the world. Startups – and established companies – are focusing massively on this technology these days. But Rina Joosten-Rabou saw the power of artificial intelligence a decade ago. Her scale-up Pera helps companies use AI to find the best applicant for a job opening. And it works. The algorithm is a better matchmaker than human HR staff.

Entrepreneurial lessons

Rina Joosten-Rabou will be the keynote speaker at the LEVEL UP event in Eindhoven on Sept. 30, 2024. She will share her entrepreneurial story as CEO of Pera. How do you build a successful company that makes an impact? How do you determine your path in a high-tech field that is changing at lightning speed? During Level Up, Joosten will share her most important lessons as an entrepreneur.
“AI is hot now, especially with the development of chatbots like ChatGPT, but it was a lot less so ten years ago,” states Joosten, co-founder and CEO of Pera. So why did she choose to deploy this technology anyway? “I believe a lot of talent is overlooked. It’s my mission to change that.” She saw that all other methodologies for selecting people for jobs were not working well. “People are often poor at identifying talent. They often base their decisions on hiring new employees on subjective information. That’s why I started looking for a new way to go through this process.”

A fairer and more accessible job application process

Pera offers AI screening that makes the application process fairer and easier. Applicants digitally answer three open-ended questions. The answers are analyzed by a set of algorithms. The system looks for more than two hundred markers that are not visible to participants. “AI not only looks at traits more objectively, it sees connections that humans don’t see. The algorithm looks at what someone writes as well as how someone writes and links this to information from successful employees in specific jobs and cultures. This makes it more likely that you will hire someone who fits the job description,” Joosten says. Pera is now used for major brands such as Rituals and L’Oréal, as well as rapid growth companies such as Heliox and Groendus in the energy transition.

Power law

The power law is essential for reflecting Pera’s impact. The impact and performance of employees in a company often follow this power law. Twenty percent of the people provide 60% of the output. The output is different for each company; it could be sales performance, profitability, or innovativeness. “If you understand what makes that twenty percent different from the other eighty percent, you can identify people who will perform best in a specific job. With that, the impact, i.e., sales or profitability, also increases. If you manage to shift one percent of employees from average performance to top performance, that has a 2.5% impact,” Joosten states.

Collecting data

To develop good algorithms, it is essential to analyze the application process. This will reveal areas for improvement, which Joosten and her team saw while setting up Pera. “Only we soon discovered that very little data was available on this. Then we started collecting data ourselves,” she says. Pera spent eight years collecting information from employees in different companies. “We asked employees to rate each other’s competencies and requested objective performance data. That gives the clearest picture; colleagues know each other well,” she says. All that data – from more than five million people by now – forms the basis of the platform. The system can compare the markers from the three application questions with the best-performing people in a particular job.

Commercialisation

Pera is now in the process of commercializing and scaling up the platform. “Now we are dealing with customers who will use the platform. They have very different questions. That also creates another way of working for us,” she says. Thus, every phase of a company has new issues, successes, and challenges. “It’s a good thing we don’t know everything in advance because then we might not have started it at all,” she reflects. Entrepreneurship is a path of trial and error. Building a business and constantly adapting it to new developments and insights.

Perfect timing

Joosten looks back on the company she is building with a good feeling. “In retrospect, our timing turned out to be perfect. AI has taken off in recent years; the computing power of computers, for example, has increased significantly. We used that time to develop our product. It would have been too late if we had to start now.”