KB 2.0 : From Data to Insights, Policy, and Impact

Teresa Chahine (TC):

Welcome to Impact and Innovation. I'm Teresa Chahine and I'm inviting you inside my classroom at Yale School of Management as we grapple with questions on social entrepreneurship and impact.

TC:

Welcome back everyone. It's episode six of Season 10, and I'm back with Khushi Baby who also kind of just celebrated a 10 year anniversary of their own. So Khushi Baby came to speak in my class seven years ago in fall 2018. And back then you were just, you were an innovative design company that had created a wearable to track childhood vaccinations in India. Correct. And that's what we talked about in my class and we thought it was so amazing and everything now, seven years later, I don't want to say completely different, but grown and evolved in so many ways and that's what we're going to talk about today. First, let's do intros. If you can just each introduce yourself and then I'll ask you about Khushi Baby and what you do as an organization.

Ruchit Nagar (RN):

Yeah, I'm the CEO and co-founder at Khushi Baby. I'm also a doctor, training in pediatric critical care and very happy to be here and happy to be here with you and with Shahnawaz, my co-founder.

Mohammed Shahnawaz (MS):

So my name is Mohammed Shahnawaz. I'm the COO, co-founder at Khushi Baby. I look after the ground operations in India partnership and happy to be back here at your podcast.

TC:

Yeah, it's just incredible that you're back all the way from India and that you got to visit twice and that we got to podcast twice and the world has changed so much. I mean it was either 2018 or 2019, I don't remember, but soon afterwards the pandemic happened and you and all your team, I think were living together in India working around the clock, and I think that was a turning point for Khushi Baby. So I'll just rewind a little bit For our audience who never heard the first episode that we recorded together, Khushi Baby started out of a class here at Yale that Ruchit was taking where they challenged everyone in the class to utilize human-centered design thinking to come up with solutions to help improve childhood vaccinations. And tell us a little bit about the early days and how you connected with your co-founder and COO and then we'll fast forward to what happened after that and tell us about what Khushi Baby did back then.

RN:

Yeah, so I mean, it was a really cool opportunity to spend a whole semester working on a single problem statement and to research the landscape, do rapid ideation and then try to refine a prototype and build an MVP. But a key principle of that class was to take our work and take it outside of the classroom and take it to the field. And funding becomes an important piece. How do you take this and make it real partners become another key ingredient of that So we had to reach out and make a business case to partners on the ground who are doing vaccine work. And we were lucky enough to win a grant here at Yale, which is still ongoing, the Thorn Prize, but we still were left with this question, okay, what do we do? We're going to go to India and try to work with an NGO on the ground, but we need to build a team. We need someone who understands the lay of the land. And so the way that I met Shahawaz was a lucky coincidence. I sent out 40 different emails to local professors hoping that somebody would reply, and one of them did reply saying that I have a doctoral student that's interested in this topic.

TC:

So for our listeners, one in 40 sounds like a low response rate, but it's not. So when you're out there trying to create change, just reach out to many people as you can and eventually someone will reply

RN:

Exactly, 10 years later,

MS:

It was like I was waiting for that meal actually, I mean at that time I was pursuing my doctoral in India and

TC:

In health management.

MS:

Yes, in health management. And the kind of data quality I receive for my thesis and for my research work is something which is always questionable. So I was looking for some action research, I mean why we should not improve the data collection process itself. And that's how this opportunity came. I met Ruchit and I mean we spent our summer of 2014 in the field and we build on the solutions that the classroom project started.

TC:

You know, Shahnawaz, this is the first time I'm hearing that story in particular, that when reached out, when your doctoral advisor forwarded you the email, you had been working on your dissertation and you were not satisfied with the quality of data that you were using on your dissertation and you were wanting to do more like action-based data or data for action, and then this opportunity just landed on your doorstep. So you quit your program to do this work, and now that's what you do for a living is help create that kind of data that was missing from the world and you weren't the only one suffering from it as a researcher, but also the government didn't have the right data to make decisions and the health workers didn't have the right data to serve the population. So wow, I never heard that origin story on your part.

I love it. One of you was out there in the field dealing with the problem. The other one was in the classroom, challenged to come up with a solution and then somehow you met in the middle. So that's where Khushi Baby was when you came to visit pre pandemic. And then when we reconnected after the pandemic, you had been faced with the question of now that the government health system wants to just pivot for a moment to focus on COVID-19, do we stay focused on our mission to improve childhood vaccinations or do we seize the opportunity to engage with the government to work on other topics, but really on the same problem, the problem you were solving, the problem you decided to solve within the challenge of improving vaccinations was the data, right? The mothers would get, because in India it's primarily mothers taking their babies and they would get this physical card from the government. The community health worker would have a logbook where she would be tracking all the vaccinations, and these are just papers piling up everywhere. And your idea was to digitize it in a wearable form and for the mothers to put around the baby's neck. I think there was also a story behind that, that you looked at all these different ways to digitize it and then you realized that in Rajasthan where you were working, they already had a black necklace

With a little red circle around the baby's neck. Let's see it.

RN:

Yeah, so it's the black thread, and then this was, so they had an amulet or some kind of attachment to the black thread, but the black thread was what was important

TC:

To ward off the evil eye

RN:

And actually common across many cultures. So we thought it was a good opportunity to tie the tradition with technology actually.

MS:

So that was a story actually. And around the pandemic, when we just before the pandemic actually, we thought that why we should not scale this across the

TC:

Beyond vaccination, beyond

MS:

Vaccination, beyond one geography. And when we approach a state, we actually encountered so many challenges in terms of the dimensions of the health programs itself. There are so many health programs are there. I mean for those who know, public health denominator is something which is very important. The target population,

TC:

The total size of the

MS:

Population, yes, the size of the population. And we have the same population where all these health workers are working, but we have different application on different registers to capture those information. So we started working on to have a unified solution, which can captures the denominator family folders in digitized way, and then you longitudinally track them with unique ID under different health programs. So that was the approach we taken just before the COVID and then we pivoted. We'd like to tell more about.

RN:

Yeah, and I mean I think we started off with a focus on vaccination and when we went to the state, really we got to see so many different perspectives and really like Shahnawaz, I said, learn the different dimensions of the problem. And then at the same time, COVID became a catalyzing moment for us where the world was saying, Hey, do you want to come and help on another big public health issue or do you want to stay in your domain, which is maternal child health? And on one hand we were already kind of seeing that the health system was more complex and it needed better linkages across the primary healthcare system. Now on the other hand, there was this opportunity that if we kind of expand our work, maybe it'll open more doors and allow us to expand our expertise and our solution in a way to a larger scale.

So we made the decision that, yes, let's go for it. Let's try to take advantage of this moment where the regulation and red tape is less, where people are already sitting together different people from different departments and can make things move quicker. And let's challenge ourselves to reimagine what the solution looks like now at scale. So this necklace is not going to be procured in the course of a few weeks. We are not going to be able to rely on procurement cycles in the middle of a pandemic. Let's think of working just on a mobile application that can run on every Android phone that doesn't require other kind of dependencies. So we had to change our thinking. We had to kind of separate ourselves even from our original baby, which was this concept that we loved to make it more practical for this new level of scale.

TC:

I guess that's why these days we've been kind of internally in the classroom and our discussions referring to Khushi Baby is KB because now it's known as KB because that's what its initial abbreviation, but also it's not just about Khushi Baby, which means happy baby. Right. It's really just about your brand as an organization that helps leverage accessible data to help make decisions and unify the health system. Correct.

RN:

Exactly.

TC:

Think one thing I heard from the conversation today was that a unique point that happened in the pandemic, like Shahnawaz was saying about social movements, sometimes you're already working on something. It wasn't that you hadn't been thinking of expanding beyond maternal child health. You already were. But then the pandemic provided this moment in time where it was just an opportunity and stakeholders were aligning in a different way. And I think that happens a lot with social movements is that people are working on something for the longest time and they just have to be ready to jump in when there's momentum and resources. And one thing that I think was aligned and during the pandemic was that you shared in class that the health department is just like a siloed part of the government, and then the information technology department is a siloed part of the government and they're not sitting in the same room talking. But during COVID-19 they were, and that allowed you to work on IT solutions for health. Now it seems like it went back to being siloed, but maybe you and others can be a catalyst to get them back in the same room. One of the problems is that the government is so siloed, even within health, it's like maternal child health, non-communicable disease, other aspects. And your challenge seems to be that you're trying to integrate all this. Correct.

RN:

And we have to figure out what is the place where it's easy to integrate and where we can kind of defer some of that control or I would say delegate some of that responsibility. So we would like to have one streamlined application. Of course there are other existing solutions and vertical solutions, and everybody nowadays is trying to make an app for the community health record.

TC:

You're not the only one. There's a whole app ecosystem

RN:

Ecosystem. So when you're in that environment, we can continue to kind of push for the best user experience, a streamlined application, but we also have to figure out maybe our leverage is not on the app side. Maybe it's more on the data side. So let the data come from whichever app the government wants. Can we be adaptable enough to take that data and turn it into insights and turn those insights into actual policy change or program change that can lead to the impact, which is what we were ultimately looking for. So no matter the healthcare domain, whether it's child vaccination or nutrition or non-communicable disease, a lot of the data that's being captured right now is just being captured for reporting purpose, but people are not making the most, or we feel that there is much more potential to use that data and to really go down and give insights down to the village level about where we can allocate resources.

TC:

So now that's what this episode is about as compared to the first one is KB 2.0. You're no longer digitizing data or generating data. There's so much data out there from so many apps and so many stakeholders and you are trying to generate insights. How do we make sense of this data? How do we synthesize it? How do we actually get it to the last mile where the decision makers are using it to change practices and policies? So what does that look like on the ground? What are some of the challenges in managing this transition to KB 2.0?

MS:

Right. So I mean we know that we have started with the data collection tool actually, and we realized that there are other applications out there actually. And in this process we took the approach to give up platform-based solution and one stop solution for the ground labor health workers. And we work a lot to help them to improve the data quality. Also, we have progress a lot in terms of having the data, in terms of having the data on the realtime basis, but it still need to do a lot in terms of data-based action no matter which source of data you are actually working on. So last decade we spent a lot on developing that platform. Moving forward, our focus is more on to give this more action oriented approach. The challenges is that at the ground level, we have the quality of the data actually. So we are using data science approach actually, and we are using AI-based approach to help the frontline workers to actually improve the data quality, give them nudges.

We have data quality score data quality metrics actually during trainings, during our interaction with them, during, we have a system where we can broadcast mass broadcast or messages and we can flag also. So all this approach is actually helping to improve the data quality. And the other aspect is that sensitization of the officials and the ground level workforce to use of the data actually. So broadly speaking, the approach is just to collect the data. Earlier they were collecting on the paper. Now the approach is just to collect on application and just to report it back where portal, whatever portal it is directed to actually. But our focus is to sensitize the officials. In fact, having a orientation station actually. So now we come out with one very innovative approach to give them a unified platform of capturing the data from the 40 different portals on the same dashboard actually. And honorable Chief Nestor launched this year only, and now they can see data from the different sources on the single platform. And then you can do the action based on those information. I mean automated letters, the government letters directly to the district levels, the chief medical officers, if they're doing good, bad or average, they should be informed. So there's a process of closing the loop of the information actually in place. So yeah, the work is on the progress

TC:

It sounds like KB has been working on this full time in terms of getting to the two point and to the next 10 years vision, but I'm putting myself in the shoes of the government workers that make these decisions and programs. And that's not what they've been working full time on is this aspect of the work, right? They're actually trying to run a government health system. So I wonder what your thoughts are about if you are given a challenge today, if you were in a totally different class that said, here's the problem statement we're going to give you, government health workers aren't using data to adapt practices and policies and the data isn't getting back to the community in terms of community health outcomes. What would you design to meet that pain point of the government health worker?

RN:

Yeah, I think apps and dashboards, the way that they've been built over the last 10 years, not a lot has changed. And now that's part of the reason why we're seeing a proliferation, almost like a junkyard of apps and dashboards. Everyone's building a dashboard. Everyone's building

TC:

A junkyard is a good word. It's just going to waste.

RN:

Yeah, it's going to waste. Some people will use it occasionally, and a lot of effort actually goes into building these systems. But at the end, I think you have to go back to this human-centered design frame and really think about who are the end users in each case and how do you make that experience more human, more magical, more useful, that they feel intrinsically motivated to use that tool and they really feel that it can help enhance their work. So if it's for health official, you need to think what are the levers that the health official can push on? What kind of resources do they have at their disposal to play with to allocate or reallocate and what kind of bandwidth they have to make those decisions? And based off that, can you give them insights that directly link into making those decisions and give them the confidence to make those decisions in a more evidence-based way for the health worker, what is the interaction in terms of how she is using the smartphone and how can that be something that can help build her capacity or make her interactions more human?

So what's really exciting now, if I were to frame the challenge is that you have a new kind of catalyzing moment coming through generative AI and these new exciting technologies, which are still in their early stages still have limitations, but is there a possibility for us to take a human-centered approach combined with the new catalyst that is coming through these new technologies to humanize the process and activate these champions, which are already within the system? So that's what I would challenge. And in fact, on that note, we are actually planning a six month long healthcare hackathon for India based off these very concepts in which we want people to have exposure to the field realities and go to the field. We want them to speak with different disciplines like whether it's policy tech, public health, data science, and we want to give them a sandbox of tools, whether it's data sets, code base or APIs, other tech tools that they can play with to kind of build for the local problem statements that they're being exposed to and really kind of make this a more community centered approach that KB is not the only one that's building these tools and knows everything about every data point, but local implementers can use those data sets to measure their interventions, or local students at the technical university can build off the code base and add an important new feature that enhances the overall user experience or connects better to the beneficiary.

So those are the types of things that we are trying to unlock through the power of the larger ecosystem.

TC:

I think you already answered my next question, which was going to be who's this hackathon for? Who are you hoping will participate? And it sounds like you're actually hoping the government health workers will come engage with this and local students.

RN:

Yeah, no, I think there is a open source community that I think the potential of that open source community is yet to be fully tapped, and we really also want that open source community to be interested in and engaged with social impact projects. Public health historically has not been an area where data has been very accessible or technology has been disruptive, technology has been quickly incorporated, but we all saw with COVID how important public health can be to every single person. And now there's an opportunity to use this technology to really affect a large population change.

TC:

It also sounds like this is an opportunity to strengthen the flow of information from the front lines to the decision makers, and potentially, I wish ideally to help the front lines be the decision makers, and I'm thinking about the parents themselves and the community health workers instead of just waiting and hoping that the data will get in the right hands and that the right hands will have the bandwidth to make the right decisions that they can somehow be involved in and inform that process. Do you see that happening at all?

RN:

Yeah. I mean, do you want to Shahnawaz speak about how important it is for us to work along with the community health workers and the beneficiaries with our field team?

TC:

To what degree and how do they have a voice in planning programs or policies that they will then be implementing?

MS:

Yeah, so unfortunately, I mean these frontline workers, the committee health workers, they don't have much say in terms of policy making and in terms of how the process should be. We actually try to play a role of a catalyst to unlock this connection with having a dedicated community engagement team where we actually see digital rollout itself a very, very, very painful process in terms of, I'm talking about few years back when we started with them using a smartphone troubleshooting and all that requires an effort. And when you have a trust based community engagement system, actually they're more ready to adopt new things when there's someone who can hear actually.

And similarly based on we have a system where we talk to beneficiaries directly. We talked to the ground level frontline workers in terms of the action based on the data, actually the malnutrition cases are there. We actually helping them to go for the referral and get the referral properly connected with them. I guess give me one more example with how they're actually excited, feeling excited. What they're actually collecting is earlier, I'm taking an example of the Tibor clause. TB is a big health issue in India and they have a process of active case finding. And in this process they just have to go each household and just to capture the symptomatic cases through our digital system, we actually give them a sorted list of vulnerable population. This actually help them to optimize the resources and focusing more on the cases which actually really need the real attention actually.

And when we have a system, when you're reporting this info testing and the positive or negative, there are high chances that you get more multi patient cases actually, and this is something which is contiguous and it's very, so this is a disease which requires a community-based approach. And so they're actually feeling now they can see the list of cases who were positive because earlier they have to dependent on the system to get that information. Now the palm top, they're actually receiving the list of beneficiary, a list of the cases, and they can focus more on the quality care. So I mean this is how, these are small examples with which can do wander at the ground level. So we are not saying that the digital space is there, it'll be there, but it's all about how smartly you are utilizing that platform. It's also about the adoption. So more trust, more committee engagement, more adoption.

TC:

So it sounds like what you're talking about is really a whole other level playing field in terms of how information flows and how it's used and training people to use it also and engage with it and being more hands-on in that way. How has your team and your organization evolved as the scope of your work has grown? What does KB 2.0 look like and how are you managing this very rapid transition?

RN:

And I think that's the burning question that we discuss every day. In fact, I think KB 1.0 was very much focused on the digital health tool rollout and getting it up to scale. So we had an interdisciplinary team and we still do a field team, a policy team and a product team, engineering team, data science team. But the large thrust was on conducting trainings, engaging health officials, going to the field, trying to get people to adopt the underlying platform. And I think we've been able over the last four or five years now, not just on to get to a certain scale. So now as we think, and that scale has even given us the confidence to go to two new states. So our team size has increased so that in each state we have at least 20, 25 people. Each state has a field team, each state has a policy team, each state has some engineering team as well.

But now as we think about KB 2.0, our goal is shifting along with where we're trying to target our efforts. So our efforts are now moving away from just deploying digital tools and more towards using data for action, using data for program strengthening, for policy change, for practice change. And this requires different skill sets. It means that we need to think a little bit about how we can do things in a more scalable way. So we won't have a trainer for every block, but what is the training of training program of the future? How can we use AI mass broadcasting or mass communication to reach people and still have a bigger impact? And we're moving away from building the software to being more of an advisor or capacity building type organization or expert in the digital health and data science realm. So our goal actually is to reduce our team size and focus it in a way that we really focus on more on that problem and can still continue to grow our impact more deeply as opposed to just growing from a scale standpoint.

TC:

It's a whole different competency and group set of roles that you'll be hiring for and learning yourselves, right? Yeah,

RN:

Monitoring and evaluation, data science, policy engagement are becoming incredibly important. Research and development. So we're still trying to keep up with all the latest technology and we do believe that they have transformative potential, but really having kind of rejigging our focus, which is our product and tech team, and turning it into a AI innovation lab which thinks about the next generation of products, if you will.

TC:

So you'll continue to have innovation and implementation in terms of community engagement and then data evaluation, monitoring and evaluation and turning that into policy. And then I'm hearing this new training capacity building just partnership work.

RN:

Exactly

TC:

Right. It's kind of a different organizational framework. And also I was thinking there are a few poverty alleviation organizations that are already doing this focusing on evidence to drive policy. So I imagine you'll be partnering with them as well or probably already are like J pal evidence action.

RN:

Absolutely. No, we work with them and in fact, they're also having to figure out what's the future of evaluation going to look like? Because the traditional approach of you do a big study and wait for two years

TC:

And it's so expensive and takes so many people and time. But now with AI, it's,

RN:

And the intervention has changed. Well, it changes dynamically and the people that you're comparing against, that whole context is being changed month to month, week to week. So we really do need systems that can evaluate in real time and think of it even more as from a tech focused angle of how do you improve the product experience by having real time monitoring of useful engagement. Not just, oh, how many minutes did somebody spend on the app or dashboard? But how much of that was actually useful towards the impact that we want to generate? How much of that was a high quality review of performance across different geographies? How many of the conversations that are happening between the public health call center and the beneficiary are of high quality and can those even be automated with the use of this technology?

MS:

There's always a gap between lab and the field actually. So we are someone who always tries to bridge this gap. And because see, rate of change of focus is different at different level. So we may be talking about the new technology, but at the ground level, the reality may be different. Maybe they're not yet ready to accept those technology. So how you use that technology in incremental way so that they actually get benefited from this process, not only around the community health workers, but also the real beneficiary, real population.

RN:

You have to have a really refined problem statement in order to not just run after the technology. And I'll give an example here. So we picked anemia as a problem. We are working in maternal and child health. Mothers have anemia. Half of India's mothers will have anemia. And right now they have an invasive test, which is costly and depends on the supply chain in order to test for you, get the test strip and put it into the portable machine. So we've been working on a new AI ML solution that takes a picture of the eye and based off that, it can estimate what your hemoglobin level is. Is it above nine or below nine? And then you can refer the mother to get referral care at an IV iron treatment at the facility, or you can just give them iron tablets based off where you are on the cutoff.

And we are picking this problem statement, not just because it's cool or the technology, the image processing is getting there, but also because we have to think about if we screen this person and this person ends up screening positive, what's the next step? Do you have something to treat it with? Or if they become a false positive, do you have a way to counsel them against that? Or if you screen them negative, what's the next step? So we have to think about the whole chain and we realize, yes, actually India is investing a lot in anemia. They already have a whole backbone for treating anemia. They actually do need help in identifying it earlier. So that way we don't wait until the end of pregnancy. It's about picking the right problem statements that fit well within the existing workflow and then kind of going deeper in there. And we don't have to build all of the AI solutions ourselves. We also believe in the platform approach. So somebody else has built a cool solution, can we plug it into what we are doing and use our network to kind of distribute that already proven result?

TC:

I think I'm kind of reframing KB in my mind as an innovation company or a design thinking company, your first design was around that problem statement, but then as you work, you're constantly presented with new problem statements, so you're constantly designing and innovating. So that's how I feel like it would make sense to brand kb. It's not just, oh, Khushi baby, that one, like wearables. It's really about finding the problem statements and finding out how you can utilize human-centered design thinking to innovate and implement new solutions.

RN:

I mean, that's why we still have a field team and a field lab. So even though we work at the state level in three states, in each state, we've made a commitment. And for right now, we're only going to stay in three states, but each state we're going to have a one district lab where we have a field team, we have an intense in the field, we're experiencing what's happening. And from those observations, like new ideas emerge and we can actually stress test whether or not the innovations make sense and fit the problem space.

TC:

I wonder if you can even find different words that could be captured by kb. Maybe it's like Sanskrit words or something where KB stands for field lab or for design lab or something like that. And just to reinvent the name.

RN:

Yeah, I think so. I

TC:

Think you were going to say something though.

MS:

Look, I mean, see, that's how you remain rooted actually, when you have a field lab in age geography to capture the diversity, stress testing. And actually, I mean, whenever our lab team starts feeling arrogance about their product, I say, just go and test this in the field. You'll get enough feedback. So that is something we keep in our mind always.

TC:

And I love that you used the word rooted. So I think your tagline is like community rooted data insights.

RN:

I like that

TC:

We're reinventing as we go.

RN:

Yeah, no, no, this is good. This is actually

TC:

And at least we're capturing it on camera, so we won't forget our ideas as we brainstorm. So what's a good note to end on? What else do you want people to know that I haven't asked you in terms of thinking about what's next? Should we come to your hackathon?

RN:

Yeah, no, come to our hackathon. We honestly, I think there's a lot of bright talent around the world and for whatever reason, I think the social impact space may not have marketed itself in the best way or made itself as presentable, but the problems are incredibly complex. Everyone on the team gets really passionate about working on these problems because they affect real lives and it requires the best of the talented resources out there to address them. So we want people to come join the social impact space, get your hands dirty, go to the field and feel meaning in the work that you do. And whether it's Khushi baby or KB or any other social impact organization, just more participation is really what we're encouraging.

TC:

Okay. That's still part of me. I can see it in your eyes too. We're still brainstorming your ahead about new names and taglines and the signs and everything. Okay. So this is going to be lots of fun. I hope to visit you out in the field in the months or years ahead and see this awesome work and support it in any way I can. And thank you so much for coming back and sharing your growth and your new challenges and your new problem statements with us. We'll stay tuned to see what you innovate and what you design next.

RN:

Thanks so much for having us. Thanks. Thank you.

TC:

I'm Teresa Chahine and you've been listening to Impact and Innovation. Subscribe to stay tuned and follow us at Teresa Chahine and SOM Ventures. Special thanks to the broadcast center at Yale School of Management.