SourceCon (a subsidiary of HR firm ERE Media) recently hosted its annual competition to crown the best sourcer in the world. In the competition, sourcers and recruiters received three separate job descriptions and titles along with thousands of actual résumés.
Armed only with that information, recruiters around the world had to determine which candidates were sourced, interviewed and ultimately hired for those positions. Among the competitors was an artificial intelligence — a machine. That machine, San Mateo, Calif.-based Brilent, placed third in the world against its human competitors in selecting winning candidates — and it only took a matter of seconds.
Brilent is an AI-based candidate ranking and assessment engine. Essentially, the technology plugs into and scours companies’ internal talent databases (ATS and CRM), cleans, updates and standardizes the data, gathers additional information on candidates, and then ranks and matches them against open jobs based on an AI-driven algorithm. The system uses additional context in its candidate ranking/selection such as job history, the employer’s historical and trending hiring preferences, and relevant roles outside the actual job description. “We started Brilent because of our experience working and hiring people internally at Facebook,” says Garry Ma, Brilent’s CEO and co-founder. “We were building products using the latest technologies in data science, machine learning and AI, but then using much more manual processes in sourcing for candidates. We thought, ‘why not apply this kind of technology to screen candidates and identify those that are a high fit?’”
Brilent is not the first candidate-ranking solution, but it is among the first to apply real AI to the problem. Ma says existing solutions are still largely based on Boolean logic or keywords, which means recruiters have to be knowledgeable about the skills associated with a role to be successful in searching. “If you’re searching for a rocket scientist, you’d need to have a pretty good understanding of what the skills and backgrounds are of a successful rocket scientist for you to be able to compose a successful Boolean search string,” Ma says. “With our solution, this is not required.”
Recruiters are the primary users of the tool. The platform returns a list of ranked candidates against open jobs to recruiters, who can view a summary of job fit, why the engine ranked candidates high or low, as well as candidate salary information and availability. Recruiters can then engage with the candidates recommended by Brilent directly from the platform. And while human-based sourcing can take hours to days for a given role, Brilent can return a list in seconds. The company charges a fee based on the number of candidates managed.
Some thought AI was nothing but science fiction. Others feared it was merely a Pandora’s Box. Brilent’s business model shows that running with the machines is not just reality, it’s a good business decision. z