How synthetic data will save the world from the Coronavirus? 


As the Coronavirus (COVID-19) spreads like wildfire across the globe without an end in sight we have to look to radical changes in order to prevent the continued spread of the virus. One way to battle the spread of the virus is to track, pinpoint and understand the epidemiologic spread of the disease. This is where technology comes front and center to the fray. At the beginning of the spread, one of the AI companies predicated the outbreak with one of their AI models through the tracking of publicly available data through social media sites.


With the Coronavirus in full acceleration, the question is - Can we deduce with technology the same insight now that the AI did at first? Some of the questions pertain to patients’ patterns of behavior, travel, interaction, flights, geographical spread, hospital visitations and more. The simple answer is ‘YES we can’ if AI companies had unfettered access to the data airlines, governments, hospital, credit card companies, and banks hold on us, with the current AI models and technology we could predict the rate of spread, where it will spread next, how and who will be affected, how long will the epidemic continue and who will be infected.

Luckily or unluckily the data held by the above industries are full of Personally Identifiable Information (PPI) about us and they cannot be shared under the current regulations across the globe. All our PPIs are protected by GDPR, HIPAA and various other regulations across the globe trying to protect us from companies looking to abuse our sensitive data.

The very burning question is how can the companies share personal data that can help fight the Coronavirus without impacting or risking our individual’s privacy? 

The answer comes in the form of synthetic data. Synthetic data in its simplest form which makes up data that fully maintains all the statistical properties and features of the real data without disclosing any of the personal data.

So how does that really work?

A Chinese airline will take all its customers' data from the beginning of January and run it through a synthetic data fabrication platform (Kymera DFSD). The platform will take the real data and turn it to a fully hyper anonymized data with none of the private information in it. In fact, it will change all the data regardless if it's personal or not. The data will hold all the statistical information and relations to other fields and will produce the same insights as to the real data. The airline will then take that syntactic data and send it to one of their AI vendors to run analysis (like locations, personal data as age, bank account, flight details,etc.)

Now take the combined info from governments, banks hospitals, and credit card companies in synthetic data form and you could have a very powerful insight on what’s happening next and wherewith the AI models. So much so, these countries and cities can prepare, understand patterns of behavior of people and start eradicating the virus.


Yoav Vilnai CEO of Kymera Labs, a synthetic data company out of Tel Aviv is calling on companies, local authorities and governments such as the Israeli Health Ministry, the World Health Organization, the EU and others to work with technology companies such as Kymera-Labs to harness the power of Synthetic data and AI and to try helping the fight against the spread of the Coronavirus globally. Yoav also adds ‘We are facing a global crisis and to prevent it we need to act as a community and think in radical ways. Our clients understand the use of synthetic data and its insights and we are encouraging everyone to harness their data trough synthetic data solution to help stop this epidemic’


Time will tell how we as a society and the arthritis tackle this issue but we must harness new technologies to unlock the potential we have in our data without compromising our privacy. 


About the Author

Ron Kopolovic is the COO of Kymera-Labs and resides in Israel after working in London with various global financial institutes.

©2019 by Proudly created with