The national congress plaza in Buenos Aires, Argentina. Image: © eskystudio/Stock.adobe.com
Israeli start-up SparkBeyond is using AI to create predictive analytics about Covid-19 restrictions in Argentina.
Like many countries around the world, the Argentine government took drastic steps last month to protect citizens from exposure to Covid-19, the disease caused by the coronavirus.
Although Argentina is in its third year of recession, president Alberto Fernández said: “The choice is to take care of the economy or take care of lives. I chose to take care of lives.”
Argentina’s response to the spread of the virus was to impose a strict stay-at-home policy which, according to Bloomberg, poses “the risk of disproportionately affecting the livelihoods of Fernández’s mostly middle-to-lower class voter base”, who rely on the informal economy to make a living.
Now, the country is considering turning to AI and predictive analysis to help make decisions about when, where and how to lift restrictions.
According to AI firm SparkBeyond, the company’s technology has been adopted by Argentina to look at how the country can allow citizens to return to work and minimise economic impact.
Predictive AI insights
SparkBeyond said that its predictive analytics technology could be potentially used to help inform the gradual lifting of coronavirus restrictions in the coming days and weeks. As of today (7 April), Argentina has 1,628 confirmed cases of Covid-19, and has reported 53 deaths.
SparkBeyond added that its data could provide “improved insight as to which industries and citizens are least exposed and least likely to pose a ‘spreading’ risk – which is considered a first step to opening up certain parts of the economy”.
The Israeli start-up, which was founded in 2013, said it is also working with governments and cities in the UK, the US, Europe, Asia and South America.
Analysing data
SparkBeyond said that the first data set being analysed by the Argentine government is focused on Córdoba and the surrounding area. “The wider roll-out will then involve accessing all positive cases and tests in the country,” it added.
The project uses data from the Argentinian ministry of health, which aggregates travel, demographic and employment data for each citizen. The start-up then integrates hundreds of external data sources to create a wider picture of the situation.
SparkBeyond said its engine uses “blindfolded analytics”, which allows the platform to analyse and create models out of sensitive data without allowing people to access private information about individuals.
“This is necessary in this case as the use of data on individual Covid-19 carriers is crucial for building accurate models – but human inspection must be avoided to ensure the sensitive information remains private,” the company added.
