Interview with Andreas D. Christiansen, CEO of Praice — Personality + Network + AI = Performance

Amazon and Facebook keep on killing it.

Voice is the fastest medium for humans to give inputs to our devices, but, as Andrew Ng said, “the fastest way for a machine to get information to you is via a screen.”

That’s why Amazon has just announced their new Echo Show equipped with a seven-inch touchscreen display that will be able to show the information requested by users straight on the screen.

Breaching the gap between humans and virtual assistants. But also Facebook is doing something interesting on this.

Indeed, its machine learning team has created a neural network that translates language up to nine times faster and more accurately than other current systems. Perhaps a step closer to virtual assistants in every language?

After this very brief review of the hottest AI news in the space, I am excited to bring you the interview we had with Andreas D. Christiansen, CEO of Praice, a Copenhagen-based startup that provides a predictive analytic tool able to forecast the performance of new applicants inside specific teams according to their personality.

As always, feedback and questions are always more than welcome. Just ping me at if you feel like talking about AI.

Could you give an introduction about yourself and about what you guys are working on at Praice?

I started Praice a little more than 3 years ago. Before I was at IBM working with big data and analytics while taking my master in in economics and business administration. 
I had the idea that the way we present ourselves would change a lot and it was going to be based on what other people think about us rather than what we think about ourselves. This was the early days of the review economy that you started to see inside the pop culture.

I didn’t see this as a 5-star scale for every person, because it would have been very limiting. That’s why I started to think about how we could express the nuances of how people stick out from the crowd in an interesting and useful way.
My idea back then was having people creating Praice profile, attached with a visualization of relationships and the answers of the people in your network about your personality, in order to use the data in an accessible and useful way.

We started as a social network leveraging a lot of viral components in creating our behavioural profiles.

Then we pivoted and started using the very valuable behavioural data we had been able to gather from the Praice users. Right now we are building a predictive recruitment tool that forecast how people will perform in large teams according to their personality.

We have gone through different stages and, overall, we have learnt that what you need is being extremely focused and create value for a specific market and segment. You need to fix a specific problem to be relevant for someone.

Solve a very big pain for a few people and start figuring out how to scale that from that target segment.

How do you use the Praice profile and your first concept inside your predictive analytics tool?

We have people in a specific team in a company create a Praice profile and then we get their performance data. Using the answers that we receive from the network of each person we are able to extract some valuable insights on their personality. Then, we combine those data with performance data and we predict for the applicants of that specific team, on the basis of their Praice profile, how they are gonna perform in terms of the percentile.

We can apply this approach because we have data showing that certain traits correlate positively with performance and other traits correlate negatively.

Of course, it is always different for each company and for each function. You need different kinds of people according to the position and the company.
This is the reason why we tailor the predictive model to the company we are working with.

How do you extract personality traits ? Do you use any traditional personality test like the MBTI?

We do not use the typical MB approach in evaluating people’s personality. Usually, a personality test would last more than 20 minutes. And there are several problems. I mean, Young people from 18 to 25 are not going to fill out a 25-minute questionnaire!
And there are problems with the standard of personality test because they all revolve around the idea that is you judging yourself. Data, in this case, is not very useful.

There are different studies showing that if you want to predict the performance of a person, it is better to gather the data around how this person is perceived by other people, rather than how they perceive themselves. If you think about it, this makes total sense.

First, people do not really know themselves very well due to human psychology. There are several blind spots and things that we do not see about ourselves but that has an impact on how other people perceive us. 
By crowdsourcing reviews, you can save a huge amount of time. It takes less than 3 minutes to create a profile and your network will have to answer to very few questions about you.

In this way, we can gather all the data we need to provide value for our customers.

Why did you decide to focus on personality? Do you see such a big relationship between employees’ output and their personality?

If we start from a personality perspective, we had a huge amount of data in terms of personality and we started to think about where personality is really a key element impacting the final performance in a job. That’s customer oriented positions. Sales, support, retail, and so on.
Basically, the element that really matters here is personality. The companies, in this case, are using a recruitment process that was focusing on everything but personality. It is really mind-boggling to see how these companies simply use a CV, maybe with video, and that’s it.

Then they use their intuition, their gut feeling. Unfortunately, intuition is not a very good predictor of how candidates are going to perform.

I am not saying you can blindly apply machine learning to get valuable insights. First of all, you need to have the right performance data and the right connection between performance and personality.
What we offer is a breach between screening solution and performance management systems which is something unique in recruitment platforms.

You are surely part of this big trend of AI disrupting HR. Is there any other field you are not working on right but you would like to explore?

We are now focusing only on selecting the right applicant. I think there are some interesting things going on especially when you can get data from social media and try to get patterns or signals from people behaviours on social media.

This could be a very good proxy to understand who you should focus on and where you need to pay more attention in terms of workforce management. Inside this area, also in terms of understanding the interactions that employees have inside the company, there is so much value that can be created and right now we are just scratching the surface. When we talk with companies, they do not really know how to define performance and how an optimal performance can be defined. Working with different clients, companies, and sectors we can cross-apply the competence that we gain between different companies and add value also in terms of which KPI should be measured.

How do you see automation and robotics impacting your vision? Many jobs will be destroyed not due to the offshoring of some of the activities, but due to their automation.

I have to say it sort of frightens me . I am very interested in the whole automation and the effects on our society. I think there are lot of jobs which people do not expect to be disrupted. When I think about some of my friends who studied to become a doctor, financial auditors or lawyers and did this to have a “safe” job and earn a good amount of money, I see there is a train coming against this kind of jobs pretty fast.

A lot of people who thought they were doing valuable knowledge work they will come to realize that their job is actually not valuable at all.

There is going to be a huge disruption in those type of jobs where people think to be safe. I don’t know what those people are going to do.

In Denmark, we have a very good social security system, but I don’t buy this thing that we have always been able to find new jobs for unemployed people. It does not take into account what we are facing here with exponential technological development. Going down this path of technological exponential growth, I think very few people will be controlling pretty much everything and data monopolies of big companies will be getting stronger and stronger in the next 10 years. This is another challenge in this scenario.

Think about giants like Facebook, Google and Amazon levering the data they have on people and taking over established sectors like banking, insurance, and also hiring. It sounds like very far off reality, but it is not.

This wave is not coming yet and we still have some time, maybe the next 10 years, to get ready for this.

How does being settled in Copenhagen helped you?

The startup ecosystem here is very good. It has been an immense help for me. I have met a lot of investors and employees that are now working with me here. In addition, there is a lot of government funding to support startup growth as well.

However, private funding is still not as developed as it is in other places. We are right now, as a company, entering into the VC game but I think also this aspect will develop further in the future.

What is your favourite book?

The Hard Things About The Hard Things is my favourite. But the best one I have read recently is Stealing Fire: How Silicon Valley, the Navy SEALs, and Maverick Scientists Are Revolutionizing the Way We Live and Work. I strongly recommend this one. It is a great book.

What is your favorite podcast, newsletter or publication?

My favorite newsletter is Exponential View, that’s just epic.
While my favourite podcasts are “Waking up”, by Sam Harris, and “Tangentally speaking” with Cristopher Ryan, that’s a very good one.

One thing you would like to tell to an upcoming entrepreneur that want to use AI to disrupt new industries?

I would say focus. We started out with a really big vision, and you need to be motivated by a big vision, but you need to go super narrow and focused when you execute.

This lesson is not an intellectual one though, it is something you have to absorb as practical knowledge through execution. I think it is the biggest takeaway of our first 3 years at Praice.