Speech Technology Magazine

 

Can Voice Analytics Help HR Find Better Candidates?

VoiceSense's AI-driven solution says it can streamline recruitment processes by automatically screening applicant interviews and objectively identifying top candidates.
Posted Oct 31, 2018
Page1 of 1
Bookmark and Share

What does voice analytics have to do with the hiring process? Up until now, not much. Today, VoiceSense, a provider of speech-based predictive analytics solutions, announced it has launched a new speech-based predictive analytics solution for human resources that would streamline recruitment processes by automatically and objectively filtering large volumes of applicants and providing a short list of the most relevant candidates for a position.

According to a press release, the solution uses applicant voices only and eliminates the need to manually evaluate hours of video and audio interviews to find appropriate candidates, which is a time-consuming process that is easily influenced by subjective factors that lead to inaccurate outcomes. “The point at which a voice sample is collected and transferred to our solution for analysis really depends on the internal processes of a recruiting firm or HR department of a company using our solution,” says Yoav Degani, CEO of VoiceSense.

Generally, customers—both recruiting firms and HR departments—are using our technology to analyze recorded video interviews and not just a voice sample,” adds Degani. Of course, recorded audio files can also be analyzed. Typically, these recorded files come directly from the recruiting firm or HR department and are sent by the user – individually or in bulk - to our solution for analysis through a dedicated API that we provide.”

Not set up for interview recording? Degani says his company can help: “We can also be directly involved in the collecting the audio recordings. In this situation, we install our software into the company’s PBX so that when the HR representative calls a candidate, the conversation is automatically recorded and forwarded to our solution for analysis.”

The technology behind the VoiceSense solution works by analyzing the prosodic parameters of a person’s speech, which are the non-content features of speech, such as intonation, pace, and emphasis. VoiceSense says its technology can assess over 200 prosodic speech parameters and builds an AI-driven personality profile of an individual’s working characteristic, such as temperament, ambition, cooperation, communication, dependability, creativity, and others.

Voice analytic solutions often raise questions and concerns about whether or not the people being recorded know what the recordings are being used for. Whether or not job applicants know they are being recorded and what the recordings are being used for, according to Degani, “depends on the privacy regulations in each country as well as the privacy policies of the recruiting firms and HR departments using our technology. In some countries, notifying candidates that the calls are recorded is sufficient. In other countries, especially those in Europe, candidates can request that all HR data is shared with them.”

With VoiceSense’s solution, human resource teams can upload individual or large volumes of candidate video and audio interview files through an API to the VoiceSense Cloud. While a human interviewer may not be able to look past outward signs that a candidate is nervous, a machine can often dig a little deeper. Degani says, “VoiceSense objectively analyzes both the behavioral tendencies and current state of mind of candidates. Our technology can be used to analyze multiple interactions and capture the candidate’s natural speech patterns over time rather than just focus on the interview to provide a more complete assessment of the candidate’s behavioral tendencies.”

But AI and algorithms have been criticized for reinforcing biases they have been tasked with eliminating. So how is VoiceSense staying proactive to make sure its analytics to fall prey to the same pitfalls?

Our validation research determines what are the behavioral tendencies that are associated with success in a specific position and how these tendencies are reflected in personal speech patterns,” says Degani. “Therefore, such biases in question are not part of our analysis process. Specifically, our analysis is prosodic, meaning it is language and content independent. As a result, the analysis of our technology is much less likely to have any biases related to gender, culture, race, background and so on, than actual human interviewers.”

The VoiceSense solution generates a specific job match score for each candidate, which are automatically sent back to the human resource database. The job match score can also include the complete working profile report of each candidate’s tendencies. Human resources teams can then create a short list of the best candidates per position for further evaluation and consideration.

VoiceSense says its solution has other uses as well, like monitoring and tracking the well-being of staff and get alerted to early signs of attrition, dissatisfaction, and other changes in an employee’s state of mind, letting managers take steps to improve retention. Managers can also use the VoiceSense solution to build more compatible teams and productive departments based on the profiles and tendencies of existing employees.

Page1 of 1