We're connecting the physical and digital worlds

Want to know more?

Meet the Team: Nikolay (MLOps)

Viqal April 19th 2022

Hear how and why an MLOps engineer works in an emotional AI startup.

Expressing emotions across cultures

Knowledge March 31st 2022

How emotions can be detected through vocal expression in different cultures.

Live in Singapore

Viqal March 30th 2022

Crossing the borders of Europe, we enter the South-East Asian market.

Meet the team: Nikolay (MLOps)

Viqal April 19th 2022

Hi Nikolay, you are an MLOps Engineer at Viqal, what do you work on?
A: "Being an MLOps Engineer is a fancy way of saying I work on everything from the moment emotional data comes in the cloud to the moment it is shown to clients. This includes data pre-processing, training, inference and evaluation of Machine Learning models, all the way to data serving and data visualisation. The cool thing is everything happens in the cloud using AWS services and Docker containers and we try to stay as flexible as possible so we can quickly introduce new insights to our clients."
Q: "Do you deal with interesting problems?"
A: "You bet! We deal with an interesting and complex problem - inferring emotion from hundreds of voice features. Moreover, we do this in real-time which requires us to have a stable data pipeline from the physical sensor on-site all the way to the data dashboard or API that clients get insights from."
Q: "What's the company culture like?"
A: "The company culture is, well, fast-paced with almost a flat hierarchy - which is awesome! If you are looking for a place which allows you to move fast and collaborate on ideas on a daily basis - you should consider Viqal! Another good part of an early start-up is that we are still improving our processes - so if you like taking the initiative and don’t want to get bogged down in never-changing process cycles, send us your CV!"
Q: "Why did you go into emotional AI?"
A: "Well, emotional AI allows for a better overall service at restaurants, hotels, etc. I have had both bad and super awesome experiences at restaurants and hotels and if there is a way to measure and improve these experiences, then that is a goal worth pursuing. It’s also a very ambitious technical challenge and working towards it is fun!"
Q: "What’s the biggest pain in the ass?"
A: "Well, the coolest thing and the biggest pain in the ass are the same thing here - the real-time collection, processing and visualisation. Making sure that all the building blocks of the extensive pipeline from the physical sensor to the data dashboard work is a technical challenge and sometimes it takes time until you find exactly where the problem originates. We are constantly improving our deployments - making them more modular and transparent, but it is for sure one of the biggest challenges still."
Q: "What will be the biggest challenge?"
A: "At this point, we are still a fairly small start-up and so scaling our technical solutions is also fairly straight-forward. It will be an interesting obstacle once we have to scale to thousands of physical sensors, all sending emotional data. It’s important to think about scaling before it becomes a problem."
Q: "How do you see the future of emotional AI?"
A: "The future of emotional AI is exciting! Being able to more accurately and continuously measure customer satisfaction means we can improve services and make sure we provide the experience customers enjoy! Smart spaces is the ultimate goal!"

Thanks Nikolay!

Link to Nikolay's LinkedIn: here

Expressing emotions across cultures

Knowledge March 31st 2022

Being at the forefront of emotional AI, we often get the question: what about different cultures? Can happiness, satisfaction or similar emotions be detected in Seoul the same way as in Lisbon? Let's talk science: vocal emotions across cutures.

Understanding emotional expression across cultures isn't exactly new, as early as 1971 researchers concluded that humans can all pretty much detect six "basic" emotions around the world: anger, fear, disgust, happiness, sadness and surprise. Most work around this emotion recognition has however been done using facial expressions, whereas the voice expression was more neglected. In recent years this gap in knowledge has taken the international research community by storm, partially due to the availability of machine learning. Many succesful attempts have been done to create at least a dual (happy - unhappy) model of vocal emotion recognition through "paralanguage", so the way we use our voice rather than the words we choose. This work has already led to commercial applications of recognizing basic emotions from vocal expressions, ranging from healthcare to callcenters.

But what about that differences between cultures?In a 2021 publication in Emotion Review, two researchers compiled all the known evidence that's ever been collected. The conclusions were that yes, we humans can detect emotions across cultures through vocal expression. The closer you are to the culture, the better you become at predicting emotions, but in general humans were able to detect not just the six "basic" emotions but an astounding total of twenty five different emotions. Luckily for us, our co-founder and CTO Daniel Formolo has also done years of scientific research in this field, his publications can be found below.

Our R&D team is always pushing the boundaries on what's possible. Additionally we team up with local experts to better understand the locations we're entering.

Link to articles: Cross-Cultural Emotion Recognition and In-Group Advantage in Vocal Expression: A Meta-Analysis - Petri Laukka, Hillary Anger Elfenbein, 2021 (sagepub.com)
Link to Daniel Formolo's research output: here

Going live in Singapore

Viqal March 30th 2022

In March 2022 we made our first move out of Europe by going live in Singapore.

Singapore as a location is known around the world as the hub to enter South-East Asia because of its international character, English speaking population and generally very outward-focused attitude. But perhaps most importantly, it's just such an amazing city that has it all. From the gardens to the rooftop bars, and the hotels to the transportation network, everything is just top-notch. As a result, the clientele of the local hospitality industry demands a standard that surpasses most places in the world. This makes it a perfect place for us to help hospitality organizations become truly guest-centric.

Our local partner is Millennium Hotels & Resorts, a globally operating hotel organization. We're very excited to start measuring guest sentiment in their locations. This fruitful partnership will surely lead to a growing success in Singapore and beyond.

Link to partner: Millennium Hotels and Resorts | Iconic hotels in amazing destinations