Frequently asked questions

Some people feel that talking to a robot is creepy. So far, how have candidates reacted to Tengai?

We've conducted some 80 interviews to date, and the vast majority of candidates describe Tengai as warm, friendly and easy-going. Most are surprised by how natural the meeting feels－despite talking to a robot. Perhaps it's the unusual situation: sitting down with and engaging in an intelligent dialogue with a robot. You might even feel that the robot is a little too human, because it's so adept at social codes. It can feel pretty surreal.

Are there only yes and no answers?

How advanced is the communication?

There's a lot of focus on providing and gathering information in recruitment, and this also applies to Tengai. It shares information in a dialogue form about the interview and how it will be conducted, and then gathers the answers. As it collects information, it understands what the candidate is saying, regardless of the number of words and sentences used. So yes, it can handle open questions and open-ended answers.

What happens when Tengai doesn't obtain a satisfactory answer? Does it understand?

If that happens, the robot informs the candidate, asks follow-up questions, and subsequently attempts to continue the conversation and the interview.

What is the greatest advantage with Tengai?

In addition to getting a more objective and structured process, we avoid the unconscious bias we all have. By doing this in the early-selection stage, we shift the subjectivity further along the process, where it is less damaging. Additionally, we're able to invite more candidates to participate in the recruitment processes' early stages, allowing for greater diversity by ensuring a better and broader selection of talent. What we can risk losing is the detail and personalization that can give a complete picture of a candidate's suitability for a position. On the other hand, Tengai is designed to be used at the beginning of a selection process, where it's advantageous to be objective and skill-focused to find the competencies needed for the job. In-depth assessment by an experienced recruiter, trained in unbiased recruitment, of a candidate's experience, potential and motivation is conducted with the talent who progress further in the process.

An unbiased robot－is it possible? How do you ensure that the robot's information-gathering is unbiased?

Tengai only records candidates' speech, which it converts into text in real time. No other variables are involved, such as a person's accent or the pitch of their voice, their looks or gender. Furthermore, we don't let Tengai know anything about the candidates; the only thing we have access to are the candidates' names and email addresses－and we don't use this information for the purpose of identifying specific candidates. That way we keep biases out of the interview. But we do plenty more, too.

What do you teach the robot? Why is it needed? And what is the human recruiter's role in this?

The robot has been developed for many years by the company Furhat Robotics. What we are doing is developing HR-tech application software, built together with Furhat on their existing OS. What we teach it is how to conduct situation-and skill-based interviews as close as possible to a human recruiter. This includes anything we do as a recruiter－like how we hum, nod our heads and ask follow-up questions. This project started in August 2018. We then established a question tree and all the social skills associated with it. In recent months, we've adjusted some of the dialogue in Swedish so it feels as natural as possible.

This is certainly a new way to recruit. How has it affected you as recruitment professionals?

We've really had to look at ourselves to understand how we conduct interviews－which movements, facial expressions, and words we use to confirm talent as humans and as candidates. We've gone to great lengths to analyze the questions that the robot should ask. It has been very important that the questions are as clear and as concrete as possible. Tengai struggles with ambiguity, so the less vagueness there is in a candidate's answer, the better Tengai is to evaluate and validate the skill sets.

You have said that it would be good for all recruiters to partake in this work. Why?

All recruiters work old-school－and old-school skills need to be questioned in a world where the tech development gets faster by the day. It's useful to evaluate yourself, the way you work on a daily basis and your interview techniques. Even if we tend to think that we do a great job for our clients and candidates, there are always things that can be better. You just need to dare to challenge yourself. We need lifelong learning, even in the recruitment industry. As we've taught Tengai social codes, we've become aware of how we behave in similar situations with people, both that we know and don't know.

How does a series of questions asked by Tengai differ from one by a human?

There's no capacity for extra words or narrative that isn't directly linked to a question that Tengai has asked. All questions are asked in exactly the same way, in the same tone and typically in the same order. That way, it's a more fair and objective interview. A hiring manager would deviate from this formula, adjusting to the candidates' responses and thereby extracting more from them, which makes the process unfair. Additionally, as a human, they are also influenced and affected by unconscious bias about the candidates, and subject to their own preconceptions and subjective interpretations. With Tengai, this is avoided.