Our aim was to make this a citizen centric platform for a non technical audience

The Jury of the 21st Voice & Data Telecom Leadership Forum (TLF) unanimously voted Cowin as a Pathbreaker technology platform of the Year 2021. This platform has touched every citizen in our country. And saved lives.

Conceptualised, built, and rolled out in record time under Dr. R. S. Sharma CEO of the National Health Authority (NHA), Cowin has no parallel in the world of technology platforms.

Beginning

Dr. Sharma joined the Indian Administrative Service (IAS) in 1978. He is a Student of Mathematics from IIT (Kanpur) where he became fascinated by the campus Mainframe computer. “The ever blinking lights and automatic readers that swallowed cards seemed straight out of a science fiction movie,” he says in his book The Making of Aadhar.

His tryst with computers and programming has continued over the last 35 years. Dr. Sharma is quintessentially a serial entrepreneur at heart within the complex machinery of the Government. In 1983 as Joint Secretary in the Department of Irrigation in Patna, he discovered a peculiar problem – transfers, postings, and promotions of Assistant Engineers was a nebulous and messy process. He created a system of record-keeping using 5-digit codes that infused clarity.

This mindset has been his strength throughout his working life. For example in the early days as CEO of the Unique Identity Authority of India (UIDAI), he had to jump in and write code to avoid delays and bypass the complex procedures in Government for onboarding vendors.

“He would wake up early to write code and put together the first enrolment client we used for our proof of concept,” recollects Nandan Nilekani, Founding Chairman, UIDAI, in his introduction.

Dr. Sharma’s projects have always broken new ground – at scale and speed – and mostly with billion + users with trillions of transactions daily. A dream for any Startup founder today.

In a conversation with Gajendra Upadhyay, Editor Voice & Data, immediately after the TLF awards, Dr. Sharma shared some interesting insights on what went into the planning and building of the Cowin platform. At short notice, in an atmosphere of utter panic – and with the learnings from Aadhar that stood in good stead for scaling up.

We all know what Cowin achieved for the country. But at this point when there was a crisis brewing, how did you manage to keep a calm mind and go about the initial planning and design for such a massive project.

There is a bit of history behind this. In the month of May 2020, when I was still at TRAI, the vaccine was still not on the horizon I had realised and then flagged the issue – that if and when the vaccination started, we would need a truly strong technology platform to fulfill the execution, delivery, monitoring and follow up including tracking of progress across the country. Transparency of the entire chain would be of utmost importance. These was just my initial thoughts and then I retired from TRAI in September 2020 and went back to farming in my village. I love this more than anything.

In the meanwhile, the Government had set up a task force tasked with the creation of a platform that became Cowin later. On the 7th of January 2021, I got a call from the very top levels of the Government to come and take over the platform and ensure the first vaccine by the 16th of January – this was the planned date of launch.

As I have spent my entire life serving the country and the Government, I heeded this call from the Prime Minister’s Office (PMO) and returned to Delhi. On the 11th of January, I started work on Cowin. As the 16th of January was an impossible target, I managed to get an additional 28 days to make this work.

Looking at the architecture of the Platform – I realised it was not perfect. I knew that without a good architecture such a massive platform would never work well. We were talking of reaching 130+ crore citizens in one go. The scale of this was mind-boggling.

So I started with some basic steps. What was the first and foremost aim? It was to make this a citizen-centric platform – non-technical audience, ordinary people who would not be able to manage complexity. It had to be without complications.

So the username and password approach was thrown out of the window. It was absolutely the wrong way for enabling registrations. This would be a non-starter.

To make it easy and seamless the easiest tool was the mobile number. We decided not to restrict the number of registrations as many people do not have access to mobiles. So we allowed multiple registrations on one mobile. Thus we made a simple registration process, using mobile and an OTP.

Then we had to plan for it to be scalable. The architecture had to be Cloud ready for expansion and rapid growth.

Finally, we decided to make the APIs completely open. We published these APIs for entrepreneurs to come and build on them. Publishing of schedules or integrating payment gateways or even downloading certificates could be enabled from any front-end tool or application. Front End was not the biggest concern for me, it was the accurate data in the backend. That was the important driver. My objective was to build this as a platform from the word go, not as an application.

We started with Aarogya Setu as the first set of data for the backend, as it already had 20 crore downloads. CoWIN was integrated into this.

Of course, we had plenty of noise about glitches and of course, there were some glitches in the beginning. For example, we started informing the pre-existing database of how and where to get the vaccine. But this did not work out - it was ignored by many.

We moved to a demand-driven model. Let citizens come and register and ask for vaccines. This was the beginning of the digital journey – which would truly start after the injection was delivered and when the certificate would be downloaded.

In all of this what would you say were the two or three learnings from your work at Aadhar that truly helped scale Cowin.

This is really interesting. Because one of the biggest learnings from Aadhar was that we could not build a monolith system. It had to be an aggregation of many smaller parts.

Let’s look at the components in Cowin. There is registration, booking of slots, identifying the nearest vaccination centers/hospitals where the vaccine would be delivered, and then the download of certificates.

These are individual components that needed to work together seamlessly. This is no different from a train journey. When a person makes a booking, you first need to know which trains are running.

In CoWIN the hospitals were the trains. We have no control over them. They have to show their location, timings, and availability – the vaccine timetable.

We, therefore, worked on the concept of loose coupling and asynchronous integration of all of these components. We were not controlling the components or the data. The loose coupling concept was learning from Aadhar and helped.

The other was the scale and sanitised data. In Aadhar when a person enrolls, we have to ensure that this data is not pre-existing. We had to do de-duplication on the person’s data. For this, anonymised data would be sent over to the deduplication centers. This involved checking the entire database against 12 different data points (10 biometrics and 2 iris scans) of the new person. This verification had to be done for millions daily against all parameters and on the continuously growing Aadhar database. This kind of thing cannot be done in an assembly line mode. One after another. The transaction numbers were in trillions daily.

This was a polynomial problem. With every new enrolment, the number of checks would keep multiplying. This learning was also applied to the Cowin platform to make it scalable faster and more accurately.

Then there was the inclusion factor, a key requirement. We developed the access in 17 languages and all scripts.

So, inclusion, scale, speed, planning for diversity, and asynchronous mode of implementation all came from Aadhar.

For the scale, you mentioned leveraging the Cloud infrastructure. How was this planned

We were aware of the problem already. So we tested the system initially to receiving up to 45,000 hits per second. This was massive. Though in reality, we ended up recording about 1000 hits per second for registrations and bookings. But this was solved by using the AWS Cloud which has been flexible and able to rise to the challenge of managing huge surges and peaks.

What is your next goal at NHA in leveraging new technologies for the Health sector?

The ultimate goal of the digital mission which was launched by the Hon. Prime Minister in September last year, is to leverage Information Technology (IT) and all allied digital technologies to deliver affordable, ubiquitous, and prompt health solutions.

For example, teleconsultation, using digital IDs like Aadhar, digital payment mechanisms, leveraging the India Stack, use of Electronic KYC, and Digital Signatures will all be integrated for the best outcomes.

There is also the digitisation of health records – which is a huge transformation. It will be beneficial for patients in ensuring our public health records are easily accessible, and verifiable for settling insurance claims and then integrating the HMIS (hospital management information systems).

Essentially, we will use all technology platforms. This will lead to huge cost and time savings for people. Before a patient visits a hospital physically – there is a range of consultations, like examination, diagnosis, analysis, and prescription, these can be solved using technology and by preventing long-distance travel.

Pathbreaker of the Year 2021

Dr. R.S. Sharma