Could Your Next Hiring Manager Be an AI System?
A fitting end to the first phase of a long and arduous job search process is an invitation to an interview. However, imagine that you come to the interview and you will be given a set of questions and an AI system will analyze your answers. While this may sound like a scene out of a sci-fi movie, many companies have started implementing a procedure called asynchronous interviews. These interviews can be done cheaply and at scale: one grocery chain in the US was gathering as many as 15,000 per day during the pandemic. Let’s first become acquainted with asynchronous interviews and then look at the role AI can play in such interviews.
What is an Asynchronous Interview?
An asynchronous interview is a new form of an interview that is done in front of a video camera, in which you, the applicant, are guided through a structured interview process, which you do on your own. In the interview, you answer (text-based) questions in front of your webcam. “Asynchronous” means that you do not do the interview at the same time as the interviewer. The interviewer is not online at the same time as you, but instead, you do the interview independently, at a time and place of your choosing.
There are many benefits of asynchronous interviews for both the candidate and the prospective employer. First of all, there is convenience. A company representative simply has to prepare a set of questions and send them to the candidate, who can answer these questions at their own convenience. It also improves the collaboration process because, with regular interviews, there’s usually one person in charge of screening the candidates, but with the asynchronous method you can have many other team members watch the videos and provide their comments and feedback.
The proponents of asynchronous interviews champion this process as fairer and less biased than human recruiters, leading to better and more diverse candidates making the cut.
How Does AI Play Into All of This?
In the previous section, we talked about how one or several company representatives would review the video answers submitted by the candidate. Now imagine that instead of people, an AI system would be responsible for watching these videos and providing the necessary feedback and insights. This starts with how a candidate presents themselves, which includes their attire, facial expressions, demeanor, and many other things.
More advanced analysis of the asynchronous interviews could be used to predict whether or not a person’s personality will be compatible with the team. Basically, the system can determine somebody’s personality from the words they say. The training data for such functionality would come from statements of actual interviews, which are reviewed by psychologists and experts. At the end of the interviews, the AI system would calculate a candidate score based on the response they gave. These automatically generated scores can then be presented to the hiring manager who then can make a decision as to whether or not to progress a candidate. Such technology helps to make the interview objective and fair, eliminating human bias, but, on the other hand, the hiring managers must not be negatively influenced by the automated scores.
Why is Data Annotation Necessary to Train Such AI Systems?
If we want the AI system to understand and evaluate many different aspects such as facial expressions, speech, attire, and anything else, we will need to train it with the right data. This is where data annotation comes into play. For example, if going to the model to recognize facial expressions, the training data will need to be annotated with keypoint landmarks like in the image below:
Data annotators would need to place keypoints along each area of interest so the AI system can recognize the variations of peoples’ noses, eyebrows, eyes, and other facial features.
Annotating text data is a bit more complex since there are more parameters such as the parts of speech, sentence structure, intent, and many other things. All of these aspects would need to be annotated for the AI system to produce an accurate candidate score and so you can hire the right candidate to fill your vacancy.
Mindy Support Provides Comprehensive Data Annotation Services
If you are looking for a service provider to annotate your images, videos, texts, and other formats, consider hiring Mindy Support to perform this work for you. We are the largest data annotation company in Eastern Europe with more than 2,000 employees in eight locations all over Ukraine and in other geographies globally. Contact us today to learn more about how we can help you.
December 22nd, 2021 Mindy News Blog
Talk to our experts about your AI/ML projectContact us