Finding the right technology talent is one of the biggest challenges facing future CIOs. They can pay thousands to recruiters that fill positions by matching resumes, or they could trawl LinkedIn to do the same, but these approaches can often be hit-and-miss affairs. Could smart software help to take CIOs beyond resume-matching and create a more nuanced approach to sourcing tech expertise?
San Francisco-based recruitment startup NetIn (formerly Godlist), hopes to do just that. Jack Yasrebi, a software engineer at Twitter before the company went public, saw an opportunity for recruiting tools that took a smarter approach when sourcing job candidates.
“The alternatives were really not that good,” he said, explaining that his service trawls multiple data sources on the web to deduce different metrics about individuals and determine how good a fit they might be for a particular job.
NetIn creates a profile for candidates by scraping information from a variety of social media and other sources. It cross-references accounts between various sites and uses tools including name, location and image recognition to determine the probability that accounts on multiple sites are owned by the same people. Yasrebi has worked hard to avoid false positives, meaning that the NetIn service would rather omit a piece of information it wasn’t sufficiently confident about than include it.
The company goes beyond simply aggregating data by scoring candidate suitability using a key metric: ‘hireability’. “We use a rating internally to determine how likely someone is to want to be hired again or to leave their job and start a new job,” Yasrebi said.
NetIn determines this using several sources. Github provides a profile setting that lets users tell others that they’re available for hire, which is a good indicator. Netin also trawls multiple sources to find out whether a person’s most recently-listed job had an ending date. Were they recently let go by their job? If the service finds that information during its automated research, it will factor that into their profile.
The software analyzes a person’s contributions to open source repositories, and notes which programming languages they are active in to help determine how experienced they are in their coding.
Taking it further
Yasrebi’s team is exploring machine learning algorithms for more advanced features such as inferring a candidate’s likelihood for a job based on past behaviour.
“We want to do other things, like determining based on historical information whether you’re likely to make a decision based on the same pattern as before,” he said. If you tend to switch jobs every three years, then this may be an indicator. The company wants to fold these kinds of inputs into its data.
One hope is that tools like this, which aim to increase the accuracy of recruitment and find the right candidate, could create better results for CIOs desperate to find the right people.
The Information and Communications Technology Council’s Labour Market Outlook 2015-2019 suggests that in companies will need to hire 182,000 more people in three years, and that Canada won’t have enough talent to cope. Computer programmers, information systems analysts, web technicians and database analysts will be in particularly high demand, its report said.