Hirize: A program, that helps improve HR-processes
Meet Esra and Eren. Together, they developed a tool that screens resumes and helps make hiring decisions based on data science and predictive modeling
Esranur Kaygin graduated in economics in the middle of a global economic crisis. In 2009, it was therefore not exactly easy for her to find a job. She started her professional career as a headhunter in an HR firm. “But I couldn’t get jobs for nearly as many people as I would have liked,” Kaygin says. “After all, I’m only one person, and my resources and the available jobs were limited.”
Since then, the Dutch woman has been grappling with the question: how can companies improve their recruiting efforts? Her startup Hirize aims to provide the answer. The problem: “Current HR programs are not able to read resumes properly,” Kaygin says.
With this question she went to Eren Yasarkurt a long time friend with a technical background together they looked in to the current offering in the HR tech space. They’ve noticed that the majority of these systems run on basic keyword searches.
When HR departments tell their programs, i.e., to look for an account manager, the program filters all the resumes that exist online — on LinkedIn, for example — for keywords like “sales” or “software”. But not all people who might fit an account manager job have used such keywords. “Instead, they may have given themselves a fancy job title online that the HR program doesn’t know or understand,” Kaygin cautions.
Another problem could be a gap in the resume. Conventional HR programs see such gaps as a kind of red flag and automatically weed out the applicant. On the other hand, being hired by companies like Google is still something that programs see as a very good thing. “At the same time, it’s obviously much easier to get a job at Google today — just because of the size of the company — than it was in, say, 2012,” Kaygin explains.
Hirize doesn’t have any of these defaults. Using Big Data Analysis and Machine Learning, the software was programmed to use common sense in recruiting, reading the whole CV, and getting an overall picture of the applicant. “From that, Hirize not only filters candidates who are the best possible fit for the job, but it also makes a prediction of how long a potential new employee will stay with the company,” Kaygin says. After all, it is ultimately the best candidates who plan to stay with the company for a long time and for whom it is worth investing in their further training.
Another problem is to be solved by Hirize: Both human and computerized recruiters previously followed a certain bias imposed on them by the data they analyzed or personal preferences: discrimination was the result. “We programmed Hirize not to include names, ethnicity, or gender in the evaluation so that everyone has an equal chance of being found and hired,” Kaygin says.
Despite the startup-process just started over half a year ago — in the fall of 2021 — Hirize will launch in the U.S. market with its beta version as early as early summer of this year. Esranur Kaygin and parts of her team will physically move there as well.
Kaygin says Hirize’s machine learning process has now reached about the level of a teenager. She says it will certainly still make mistakes, but will also be able to learn from them quickly: “It’s important to us that we explain the product personally to our customers in the U.S. and also get feedback from them face-to-face. That way, we can weed out Hirize’s mistakes within a very short time.” And in this way, hopefully, enable more businesses to find the staff they want.
Follow Hirize’s journey on LinkedIn.