From solving hard business challenges with IBM’s emerging markets strategy division to tackling complex civic problems as the Sub-Divisional Officer (Civil), Rangia, Kamrup District, Assam, Kailash Karthik’s career is a study in contrasts. As a young bureaucrat, Karthik says he was sold on the power of data and technology in aiding efficient decision making. But his journey so far has thrown up surprising allies — those who had never used technology before.
Below are edited excerpts from an interview, where Karthik spoke to us about how data-driven decision making is helping administrators improve hygiene and sanitation, curb corruption, and prepare better for natural disasters.
Q: When did you first get interested in data-driven decision making?
It was during my time at IBM before I joined the government. At IBM, I was working on emerging markets strategy, which included markets outside of the U.S., Western Europe, and Japan. This was during 2010-11. This was also the time when IBM was making the transition to new technologies like the cloud, mobility, and cognitive computing. My [understanding of] how to leverage data and technology to drive decisions comes from this stint.
Being part of the private sector, I saw the role data was playing in all vital decisions, from allocating resources to identifying new markets. You could say that at times there was an overkill of data (smiles). Globally at that time data analytics was nascent. But one could still see that the right data in the right form at the right time could have a major impact.
By the time I came to the government and took charge here in Assam, I was already sold on the power of data. Nobody had to sell it to me.
So by the time I came to the government and took charge here in Assam, I was already sold on the power of data. Nobody had to sell it to me. Now every day one feels just how important access to data is for improving delivery and drafting better policies.
Q: What differences do you see in how the private sector and the government engage with data?
In my limited experience I have felt a huge difference. If you see within the private sector, the new-age technology startups are far more comfortable using data to drive their strategy. I don’t think data thinking is ingrained to that extent in larger companies, though they have made a start.
The government too has a lot of catching up to do. One of the reasons is that the government still operates in silos. It has a departmental approach as opposed to the nimbleness of a startup. It is a behemoth; it takes its own time.
The prime minister himself has been championing data-driven governance.
But there are encouraging signs. Take the national open data policy [which is in the works], making India one of the first developing countries in the world to have such a policy. The prime minister himself has been championing data-driven governance. If you look at something like MyGov, it has a strong data angle. I think the nimbleness too will eventually come.
Q: Could you give us a few examples of how data has helped you in your own work in Kamrup?
My journey with the government is still in its early days. That said, I have been able to carry out some experiments. One was during the election in Assam (in April 2016). One of the administration’s tasks during the election is to make sure that there are no violations of the model code of conduct. This time we launched a mobile app called Samarth, where the idea was to involve the public in reporting code of conduct violations. They could take pictures of any kind of violation happening in their neighborhoods and upload them on to the app. The GPS coordinates of the location would be captured, a notification would be sent to the ground team in that location, and violations would be stopped. The UI was simple, but at the backend we were also trying to monitor the areas from where repeat violations were being reported. We would then overlay these hotspots on a Google Map, which would help us identify vulnerable areas on a real-time basis so that we could deploy resources [pre-emptively].
We would overlay hotspots on a Google Map, which would help us identify vulnerable areas on a real-time basis so that we could deploy resources.
The results were fairly good, even though the app was launched close to the election and we couldn’t do too much advocacy to promote the app. Rangia (one of the two sub-divisions that make up Kamrup district; Kamrup Sadar is the other) is not a very tech-savvy area — people are still not that comfortable using smartphones. But we were able to showcase Samarth as a test case to the Election Commission of India. The idea was for the ECI to use it as a model, refine it, and roll it out in other parts of the country where tech literacy is higher. The ECI representatives that I presented the app to were very excited. I think right now they’d be assessing the viability of the app along with all the other electoral innovations that other districts across the country would have fielded.
The other important application of data-driven governance is the work we are doing with your team at SocialCops on the Swachh Bharat Mission. In 2011-12, a survey was conducted across India to identify villages that had open defecation. After the survey, it was discovered that there were wide discrepancies in the data. In some places open defecation was underreported, and in others it was overreported. Our district collector wanted to do a fresh survey to understand the real picture. We have identified some 35-odd villages that can become open-defecation-free by the end of the year. Results from this survey will help us get there.
Significantly, all such surveys till now were pen and paper. This is the first time a survey will be done on mobile devices.
Q: You spoke about involving citizens in reporting code of conduct violations. Was it difficult to create public awareness of these violations?
Not really. Most violations are intuitive — like bribery, distributing free liquor, using a public space or a religious place for political campaigning, etc. Secondly, even if the general public is not aware of what constitutes a violation, the political party workers are. Our objective was that even if the public per se don’t use the app, the party workers can use it to keep a tab on each other.
Q: How about the Swachh Bharat survey? What kind of response have you got for it?
The survey has been extremely well received by all the partners, including those that are constructing the toilets.
Mobile-based surveys save time and energy. You don’t have to worry about transferring the data from paper to a digital database. More importantly, with pen and paper, you are never sure if the surveyor has really gone to the field to collect the data or if he is sitting in a room and ticking boxes.
Another important benefit is, the survey will capture the GPS coordinates of all the places where toilets need to be built.
Another important benefit is, the survey will capture the GPS coordinates of all the places where toilets need to be built. After the construction is over, we will validate whether the toilets have actually been built. Often there are allegations that funds for such schemes are diverted, and there was no sure-shot way of finding out the truth. Now verification will become much easier. Of course we have to find other means to check the quality of the toilets.
Also, doing a survey and building toilets is the easy part. The real challenge is to convince people to use those toilets. Thankfully, in Assam there is growing public awareness of sanitation and hygiene.
Q: Assam recently faced one of the worst floods in its history. Natural calamities are a regular problem in this region. Is there scope to use technology to prepare better for such events?
That’s a very relevant question. One of my colleagues has been working on an app through which the government can send notifications of disaster warnings to the citizens. Citizens can also use it to send SoS alerts in the event of disasters.
It is very difficult to make logical sense of existing disaster management data or take pre-emptive action based on it.
Generally the data related to natural disasters sits in a lot of different reports that don’t talk to each other. It is very difficult to make logical sense of this data or take pre-emptive action based on it. Having a digital repository helps overcome those challenges.
The second aspect where technology comes in is the post-disaster relief and rehabilitation work. In the era of pen-and-paper record keeping, this part too faced the same problems with the credibility of data — the resources required to convert the data into digital records, records getting lost, etc. These problems don’t exist on mobile. This is a major area where companies like SocialCops can partner with the government to build systems and protocols.
We’ve helped over 150 partners collect over 100 million data points through our mobile data collection app Collect. Collect all sorts of data, including images and geo-coordinates, under the toughest field conditions with low battery and no internet.