In March this year, it was said help was sought from American software company Sprinklr Inc. to save Kerala from the threat of extermination.
A Kerala-specific study, done by a group of experts associated with the Centre for Disease Dynamics, Economics and Policy, John Hopkins University and Princeton University, predicted a devastating surge. It said about 80 lakh people in Kerala would be infected with Sars-CoV-2 between March 28 and April 25 (Kerala government mentioned this in its affidavit in the Kerala High Court in April this year) .
Seemingly in sudden panic, and also because the government felt it did not have the in-house capability to make sense of such an avalanche of data, established norms of awarding contracts were kicked aside and Sprinklr was asked to come in quick with what looked like a direct call.
Five months later, when confirmed cases are still short of 80,000, leave alone 80 lakh, Sprinklr's magic software application has turned out to be the most unwanted weapon in Kerala's anti-COVID arsenal.
"I gather that Sprinklr's services were required when unstructured data came in large quantities. Now we get structured data using conventional tools," said Dr S Chithra, the executive director of Information Kerala Mission.
The Sprinklr software is still with Kerala but has been left untouched.
In the early days of COVID it was argued that disciplining the crowd of unstructured data scattered all over the social media space - on Twitter, Facebook, blogs, emails, chats, discussion boards and sms messages - held the key to virus control. It was said that major epidemic outbreaks in the west were first detected by the World Health Organization (WHO) through these informal sources.
"Theoretically, this was promising. The concept had also worked in Europe, like in containing seasonal flu,” said Arun, a free software proponent. “It would perhaps be normal for people in the West to announce their health status on social media platforms like Twitter or Facebook. But I am surprised anyone believed that Malayalis would discuss their COVID symptoms, mostly running nose or sore throat or cough, on social media platforms in such a sustained manner for analysts to make out a pattern," he added.
In essence, no COVID secrets were hidden in scrambled social media data.
"We still have the Sprinklr software and no one has prevented the government from using it. If it has not been used, it is only because the software is unfit for the purpose. What the Sprinklr software does is manage the social media space to benefit its customers. It cannot tackle a disease outbreak," said IT expert Joseph C Mathew.
Kerala's needs are now adequately met through the Covid-19 Jagratha portal (https://covid19jagratha.kerala.nic.in/), a one-stop platform for real-time COVID-19 surveillance developed locally by a team led by Kozhikode collector Seeram Sambasiva Rao. It was with such talents like Rao on its payroll that the government went to the High Court in April and pleaded that it did not have the expertise.
Covid-19 Jagratha was originally developed to keep track of non-residents returning to Kozhikode by air and road, and also to help in contact tracing when returnees test positive. It was found so effective that it was adopted for the whole of Kerala.
One-stop COVID-control platform
It is a dynamic portal with up-to-date reliable information on who is infected, and where, and also their primary/secondary contacts. Those in quarantine, and recoveries, are also mapped. There is comprehensive hospital data, too; location of Covid Care Centres, hospitals and COVID First-Line Treatment Centres and how many of the beds, ICUs and ventilators are in use. Even the availability of ambulances can be tracked.
Such details are provided right down to the panchayat/ward level, and as a consequence the disease progression at the lowest level could be witnessed. In fact, the information is collected the conventional way, from the primary and community health centres where the testing is done, and also local bodies.
Health officials, unlike the general public, are given access to more personal details like name and address of positive cases and contacts. "We collect just the bare personal information. Even occupation of the individual is not asked for," a top health official said.
Did Sprinklr get too personal?
On the other hand, Sprinklr wanted access to a wider pool of data. The company had asked for information on international travellers, domestic travellers, health workers or people who have contact with patients, and vulnerable people, especially the elderly and those with serious ailments.
Apart from these, the government had also agreed to part with information regarding medication for even illnesses like blood pressure and diabetes.
Some such information was handed over to Sprinklr during the last week of March and early April but the High Court asked the government to quickly anonymise the data. This directive was issued on the basis of the Centre's submission in the Court that data could be shared only after it was anonymised.
Later, the government told the Court that Sprinklr would not handle any data and that it had been asked to destroy the information handed over.
Initially, the data collected were uploaded in Sprinklr's Amazon cloud server in Mumbai. But when this stoked a major controversy, the data was shifted to C-DIT's cloud server after augmenting its capacity for Rs 1.20 crore.
The Centre, too, had objected to the Sprinklr deal. In an affidavit submitted to the High Court, the Union IT department had questioned the Pinarayi government's decision to overlook several government-owned and government-controlled entities like the C-DIT and Information Kerala Mission.