Chennai-based start-up PiQube helps companies hire faster and better by matching job needs to the right candidates.
A few months ago, a Singapore-based bank was looking for candidates with knowledge of FORTRAN, a programming language that’s not in vogue anymore. Its search was desperate, as all its systems still run on FORTRAN. Yet, none of its attempts bore fruit.
When it did finally bear fruit, it was thanks to a proprietary algorithm (say, a formula) built by a Chennai-based start-up called PiQube. Started by 32-year-old Jayadev Mahalingam, PiQube wants to help companies hire faster and better, by matching job needs to the right candidates.
The start-up, groundwork for which began in 2014, was in the news a few days ago when a human resource focussed private investment company called the HR Fund invested $500,000 (about Rs 3.2 crore) into it. It was the second public recognition for PiQube.
In December last, the company and its team of 14 employees moved into an incubation centre run by PayPal, an eBay company, and The Indus Entrepreneurs network. By being one of the few to be selected thus, it not only got itself space to run its operations for a year but also an opportunity to be mentored by pros in the field. The starting point, Mr. Mahalingam says, is a database of 76 million candidates, which is increasing every month. His algorithm uses 72 different data points to zero down on the best resume. Some of these points are domain skills, mobility, the probability of this person choosing the job, social profile, and so on.
“Our algorithm mimics the brain function of a human being. As a human being, you are going to search across multiple portals. Our system has the ability to do it by itself,” he says.
The idea for this start-up struck Mr. Mahalingam when he was running a recruitment firm. He had to figure out two things — how to pick up professional information and funnel down to the best set of people. “The end game is replacing a recruiter,” he says.
He also realised that the world of recruitment was lagging behind in the use of data, unlike many other professions. But making algorithms work isn’t that easy. For instance, he found people write ‘Ernst & Young’ in 173 different ways! His point: a lot of cleaning up of data is needed.
PiQube currently has four IT firms as its clients – Mr. Mahalingam doesn’t want to name them. His business model is to charge companies for use of the algorithm-based tool. Backed by fresh infusion of funds, he wants to increase the database by ten times in the next two years. He is also looking to build his marketing team and eventually look to expand into the U.S.
Mr. Mahalingam wants to sell the company after building it to scale in the next 4-5 years.
Meanwhile, he is eager to solve more recruitment problems, like the one he did for the Singapore bank. Its FORTRAN puzzle was cracked after PiQube’s search showed most of such experts weren’t working anywhere but doing research work.