Connected: Tool adds AI to process of picking job recruits

Share with others:

Print Email Read Later

Almost unanimously, when I speak with entrepreneurs and executives about the most challenging part of building great companies, they put finding the right talent at the top of their lists. Getting the right people into key positions can explode the possibilities for your company; getting the wrong people can damage your ability to thrive.

Traditionally, we relied on resumes and interviews. In more recent years, we posted jobs on Monster or other job boards and hoped the right people would apply -- then manually reviewed the resumes from the respondents or have recruiters and experts do it. But it's still hard.

Ashish Kumar and Jaime Carbonell of Talencea ( have developed a tool to use artificial intelligence, or AI, to increase the odds of making the right choice while you're recruiting and considering candidates to hire.

When in its infancy, AI was only used for mathematical and probabilistic problems -- like IBM's chess-playing computer, Big Blue, where there are a finite (albeit large) number of possibilities and the computer can use math on each turn to determine which moves will increase the odds of winning.

Being able to break out of that mode and into a mode where AI could actually analyze unstructured answers out of potentially unlimited possibilities is where the breakthrough came to create IBM's Watson, the computer that beat Brad Rutter and Ken Jennings, the biggest money winner and biggest game winner in the history of "Jeopardy."

An ex-IBM executive, Mr. Kumar likens his use of AI to IBM's Watson, because matching the right candidate requires inferences that might not be explicitly stated on a resume or a job description.

As his example, he describes how a keyword search might get you lots of resumes, but still not the best people. For instance, take a search for somebody who can program in the Java programming language. Using Google or Monster for your search, you'd look up the keyword "Java" for matches. Perhaps you can narrow your selection to some number of people who show Java on their resume.

It's possible that the best candidate available doesn't list Java on his resume (for whatever reason); yet that candidate fits your needs on so many other levels and can learn Java quickly.

That's akin to finding great athletes who can learn the game, instead of hiring mediocre athletes who know the game -- much the way great soccer players can become great football kickers. Using an AI search, you might find those people who are the best candidates but don't make the keyword list.

Using Talencea's AI techniques allows the recruiter to scour a larger number of candidates in a reasonable time frame and provide what might be a better candidate pool for further review. It should significantly increase the odds of success when trying to find key talent. However, it takes a learning curve to understand the type of position, company culture and candidate for the job.


Follow David Radin on Twitter @dradin, or learn more at


Create a free PG account.
Already have an account?