The current crisis following the coronavirus is unprecedented. It’s touching every aspect of society and the effects can be felt everywhere. To get through this, we could use some smart solutions. These Amsterdam-based startups are employing their artificial intelligence to help others weather the storm.
AI for the COVID-19 app
Currently one of the most talked-about technologies involving COVID-19 are mobile contact tracing apps. To track the spread of the virus, governments all over the world are implementing or considering using smartphones to track infections. Upon hearing this, Manoj Adithya, founder and CEO of Amsterdam-based Geospark, knew he had something to help: “When we first heard about COVID-19, we immediately thought that our location technology could help. We knew that the government and medical institutions would need data to help identify and prevent the spread of the virus.”
Geospark offers an AI-driven location SDK and API for software builders to use. Smart use of location based information is at the heart of contact tracing, is what Adithya figured. So Geospark built an app to do exactly that: “our C19 app could give real-time crowdsourced and verified data that could help prevent the spread of the virus, control lockdown measures and even allow some parts of the country to stay open.” Getting such an app approved for official use is not an easy task, as Adithya found out. Geospark has been in contact with the Dutch, French and Canadian governments, but the approval process is slow, he says. However, their AI is being put to good use by others: “Over a two week period we had a significant amount of people create new accounts for coronavirus tracking related projects.”
Beating the tourism slump through AI
Where location-based intelligence is in high demand, some other sectors are hit hard. With everyone staying at home, tourism and leisure has come to a near complete standstill. Convious, from Amsterdam, is operating in that field. They offer a cloud-based reservation tool that uses AI to offer dynamic pricing, spread visitors and predict crowds. Just as the venues that use their software, Convious got hit hard, says CEO and founder Camiel Kraan: “The moment the virus hit Western Europe, we got hit just as bad.” But Kraan also spotted an opportunity. With a 6-feet-society in the works, venues would have to completely rethink their crowd control. Something his software was made to do.
So they launched Convious Crowd Control, to help venues in the leisure industry manage their crowds to keep a distance post-COVID. “It’s not so much different from our existing product as it is more of an evolution. We have always chosen a data-driven approach, using scarcity and fluctuations of demand and supply to help our customers optimize their bottom line. Our algorithms are based on these principles. This new reality makes the need, use, and benefits more evident to everyone.”
Even though his company got hit hard initially, Kraan did manage to see opportunities in the volatile market: “we also clearly saw the opportunity and stepped up our sales and development efforts. Our industry is not online by design, let alone data-driven. The new reality we currently face is clearly accelerating their need for our solution to get to this point faster and stay in business. In the last four weeks, we have doubled our customer base across Europe, so we’re confident we will come out stronger than ever before.”
AI serving food delivery
Similar innovation is happening at Dashmote, which uses AI to retrieve market insights from online images and text. Combined with their own AI and an easy to use dashboard, this gives some of the worlds biggest companies important information on their clients and the markets. Unfortunately for Dashmote, many of these clients are operating in the food and beverages sector. With restaurants and bars forced to close all over the world, this could’ve spelled trouble, says Dennis Tan, CEO and co-founder of Dashmote. “We had a good look to identify how much of our market disappeared and which opportunities arose.”
Stay up-to-date: Read all our COVID-19 coverage here
They soon landed on the delivery market, which is booming with everyone sitting at home. Their new Meal Delivery Radar offers detailed market insights to companies bringing food to people’s doorstep. “We have a large team, and we are able to use certain subsets of the data and parts of the UI for this new product. But it’s definitely a small pivot of what we were doing. In a normal situation, we wouldn’t have launched this so fast. During these times, however, you sometimes have to rely on your gut feeling a bit more.”
Employing AI to search for cures
But nobody is more well-positioned to handle this crisis, then startups employing smart solutions for medical research. Because if there is anything the world needs right now, it’s a quick way to get a vaccine. Castor EDC has created a platform that accelerates scientific research by capturing standardized medical data and helps scientists to make data-driven decisions. When it comes to COVID-19 research, they’ve decided to offer their platform for free.
“Among them is the large COVID-PREDICT study, currently conducted by Amsterdam UMC and Maastricht UMC+. They designed a system using machine learning to determine likely outcomes for individual patients. That is to say, who is at greater risk of requiring intensive care among those hospitalised.” John Ambrose of Castor said previously to SiliconCanals. Recently, Castor also announced they would receive pro bono support from Tableau Software, Salesforce, Alteryx, True B.V. and The Information Lab Netherlands. Castor CEO Derk Arts went as far as calling the nationwide collaboration ‘the blueprint for the trial of the future’.
COVID-19 imaging AI
Similarly, Aidence took the initiative to deploy their AI-chops in search of a solution. The startup from Amsterdam usually offers their AI to radiologists as an application to improve reporting in the treatment of lung cancer. When COVID-19 struck, they set up ICOVAI, a consortium for COVID-19 imaging AI. It consists of industry experts to develop and deploy a not-for-profit cloud-based AI solution for COVID-19 on chest CT for any European hospital to use. “The AI model will return a score. This indicates how likely it is that the person has COVID-19, and provides a measure of how much of the lung is affected,” says Mark-Jan Harte, CEO of Aidence. “We are further looking if AI could predict the disease evolution. For example, by advising on the need for ICU care. Using CT’s to detect COVID-19 is not advised by medical societies, but it is a choice many hospitals make because swab tests take too long, are unreliable, or simply unavailable.”
However a switch like that is not easily made. “From our 5-year experience building and deploying certified AI solutions for chest CT’s, we learned what it takes to build a tool responsibly and with respect to quality, regulatory and ethics standards,” says Harte. Battling COVID-19 is a process that involves doing a lot of things at the same time for the startup. “The design of our roadmap is linear. The actual process is not.” Aidence is currently building a diverse database of CT scans, began modelling, has decided on the regulatory path to follow and started preparing the technical documentation on the intended use of the solution, according to Harte: “There still is loads of work to do, but we have a strong team, who feels highly motivated and energized to contribute to the fight against COVID-19.”