Vaissnavi Shukl
India’s labor economy is heavily on a migrant workforce that travels from its rural to urban areas in search for employment. Today, our focus is on a tech platform that provides potential migrants with employment housing by minimizing the risks for the workers without compromising on their choice. Bandhu matches employers with migrant workers to occupants looking for informal affordable housing on rent. With the support of the Graham Foundation for Advanced Studies in the Fine Arts, we speak to Bandhu’s Founders, Rushil and Jacob about how they are on a mission to transform the unorganized labor sector in India.

I am Vaissnavi Shukl, and this is Architecture Off-Centre. A podcast where we highlight contemporary discourses that shape the built environment but do not occupy the center stage in our daily lives. We speak to radical designers, thinkers and change makers who are deeply engaged in redefining the way we live and interact with the world around us.

Vaissnavi Shukl
So Rushil and Jacob, I think, kind of a broad overview of what migration in India looks like, there is a certain, if you could talk a little bit about both migration and how labor moves across India for work.

Rushil Palavajjhala
It is a sort of break it up into, I mean, if you see the census, the last time you had a census, it said the most, the most common reason was marriage. So women moving after they got married was one. But it’s kind of a discount and simply try and understand this phenomenon of groups of young boys, typically boys moving to cities. Agriculture off-season, find a higher paying weight job, and then kind of come back to help out with the farm, participating in agriculture, the second headline, how many crops a year are they getting? It also, very often, determines what kind of cultivation you will be there at a certain time. You know, you start using those of us who are familiar with construction, especially in Ahmedabad, a lot of your labor that comes from the whole side, they will participate in a lot of digging, a lot of the filling of the slab, the concrete slabs. And then just before we’ll go back tight so that there’s a lot of these kinds of titles. So now to just spread, uh, explain this a little more broadly, what we’ve observed is single male migrants, typically age 16 to 2021 keep moving back and forth between religious cities. For this comes in age where around 2021, they are married. At that point, the wife and so stays back in the village, and the male migrant continues to move back and forth. But there comes a point when a migrant has adequately, adequately progressed in terms of skill and wage, and at that point they can afford to bring the family to the city, because bringing the family to the city effectively means you have to pay rent for the entire room all by yourself. Until then, five or six boys sharing the room, the rent saves 3000 rupees, split by people. Now suddenly you’re paying that three or 4000 rupees all by yourself. And there’s always, obviously integration, kind of a timeline for, let’s say, the wife to start becoming part of the urban workforce, get familiar, etc. And we’ve seen, typically, families of this sort try and become part of the formal workforce and effectively move towards more dominant or higher paying jobs. And they’re able to secure this. And by the time their 30s are, they pretty much become urban. That does not happen as families move back to the village, and then they kind of live that life. And this also is stuck to how many years can they really do? And 35 is pretty much the kind of gap of age. So yes, migration has these progressions, and people try to climb. And a very conservative may be the at least 200 to 300 million people every year in India who are doing this, moving back and forth. Estimates say, of this about 7075 crore construction. And then, of course, there are lots of other kinds of things, like sugar area harvesting and other higher wage agri and manufacturing kind of such the broad overall landscape we looked at.

Vaissnavi Shukl
Jacob, you and Rushil met each other when both in grad school at MIT, and that’s where you curated Pandora. You pitched it. Do you want to talk a little bit about how you both came together, and what the initial idea with one two was, when you started pitching it back in, was it 2019 2020? Right? Right before covid.

Jacob Kohn
I remember this is around 2019 so I guess it kind of came together as a fortunate that we kind of have interests. So I’ve been working on my thesis at MIT, which is based around studying informal recycling networks. So I was always interested in GIS coding, explaining that data in a way that you know actually you could get data from areas didn’t previously have it and use it to help vulnerable groups to implement their situations socially to India, Obviously, that was another reason to come here with a local startup, helped them put together a business case for how they could help by local informal recycling businesses to expand and become more profitable. So that was really interesting, and gave me a lot of insight into how India works, and also how these micro economies also sustain themselves, and this is around our career at MIT, where Rachel had also done his thesis. Can talk a little bit more about that, but yeah, at the end of our graduate experience, we’re both kind of looking at things to do that were maybe in the list little bit less conventional sense I could have probably joined a software engineering firm or done some kind of record based around js, but it’s gonna be a little more predictable. So let’s try this. I think probably.

Rushil Palavajjhala
Jacob the development class, we did objective corridor and upon large development issues based on this interest in what was happening with the agri economy in India, prime to agri distress, and ended up doing research in the Ahmedabad-Mumbai corridor, which was again, trying to understand rural to urban, etc. And I think part of that, I mean, that’s pretty much when the idea occurred. Different teammates. We made our first pitch, September 2018, so that was when I kept finding each other. There was another class on development, which was based in Mexico. We were working on different aspects, but again, monetization of real estate, land transport. So, yeah, you know, complementary interests around and of course, while I focused on going to the tech for real estate and CGS and data science piece happened a little bit later. And I think when I also found mutual but also complementary interests and Pando seemed like a very exciting solution to India was a data scarce environment, and somehow we kind of had our grasp on what was happening, also to pre-rate that, you know, I had been working on affordable things for a long time. 2013-14, onwards, I was working to implement the federal Affordable Housing Policy project, and, of course, with politicians, senior bureaucrats, helping them fund from the center, and in the process realizing, and of course, the lack of capacity level. You know, municipal governments are very, very understaffed. And, you know, below capacity in terms of professionals identifying and managing these kinds of solutions. So I think given all of that, it seemed that technology solving for information asymmetry was perhaps the best thing. And of course, it was coincidental with India’s digitization, with data rates falling, etc. So I think a lot of various things came together. It wasn’t just the trouble, it was also the system that necessitated attempting such a solution.

Jacob Kohn
Just to add to that, I would also give MIT a lot of credit for sales on a lot of ideas very quickly. Let’s try out and provide some ideas that’s really to their credit for being in that space.

Vaissnavi Shukl
So both of you actually came into it with something that you were both working on, which actually wasn’t what Bandhu started with solutions. You came with a fair bit of background in housing and private equity, and Jacob came into it with this background, what Bombay is, and what how Bombay works. Bandhu actually started by looking at the jobs market for migrant labor, and it is something that you then do as your starting point, and then transition into the housing first iteration of Bandhu, where you were solely focusing on the jobs market. And then, if you could talk about that transition from the migrant labor job market to the migrant labor housing market.

Rushil Palavajjhala
Maybe just a small correction there. We started off with bundling jobs and housing. That was before graduating was this course with Sid Yog at Harvard Business School. So they took all these New Age tech enterprises in developing countries and tried to solve some kind of wicked real estate problems. So yes, we said vicious because, you know, people typically have struggled to build a two sided marketplace for whatever reason. Thought, Let’s build a three sided market because there was value in the bundling. Now. So if you are a real estate nerd, if you see the big red curve, there’s land prices, there is how much you earn, there’s also transit costs. And I think while we’re at the MIT Incubator, we realized, where do we start trying to increase the marketplace in one go? And the most obvious reason to begin with seems like the job space. Because if you ask someone by migrating, they’ll sit for a job for a higher wage. People are already migrating, migrating relatives they had in the step to find a job. And then we thought, okay, let’s focus on jobs. Also, we thought that’d be something that we more monetize in the beginning, and yeah, so that’s kind of where we started on jobs, and very soon we started, is that the industry now, this is six months into our product launch. Covid has struck. And, yeah, everything is under lockdown. Migrants are and of course, industries are struggling. At the end of the day, they were still in manufacturing. Things that were still basic necessities, being manufactured with these broken supply chains of labor. And this is why we got and, you know, found new tried to close some of those gaps, and some things seem to be picking up well for a year, year and a half, and then, as the lockdown waned, as the pandemic started to ease up, we realised no real sustain from industry to digitize the work. The more time, the more workers they literally are showing proof of employment and their compliance burden was increased, right, right from DFDs, IC, etc, tons of compliances. And a rough calculation for employers showed that the compliance cost could almost shave off 30% of their top line. And this is where the lies are: how should we push into it? It is very ugly. We’ll be able to formulate this work. With regulation it is loaded against industry. If they integrate, everyone is trying to shave, you know, shame the money of whatever they can. Commodity prices were increasing. Labor crop costs, of course, are quite high. So that was where we decided that maybe the housing piece used to start experimenting with parallely and yeah, coincidentally, demand for housing was very rapid. Migrants were coming back to the cities to find employment, and at the same time, we were moving parts. Your work site is contractors, keep changing, moving. Your workers keep moving. So there are multiple at the same location in the same arrangement all the time. And housing also seems to offer fewer moving parts when the house doesn’t move, the landlord moving part really is the worker and simpler to solve and monetize. So that’s kind of where the transition happened.

Vaissnavi Shukl
Jacob, if you can talk about what it would you say it’s an app? Would you say it’s a technology, it’s an interface platform, and how does one use it?

Jacob Kohn
Great question. I think both Rushil and I have always thought about it as a holistic channel, but at the end of the day, workers get whatever they need access to. And so that’s really how we envisioned it from the very beginning, that you can’t have a job. You can’t have a job without housing, and there needs to be a central data story, basically, which is able to build out that data and understand what a worker really is. For example, when we onboard a worker, what their origin, what their sense of maybe where they’re migrating, where they’re interested in migrating, what kind of housing were based on Different kinds of demographics that were coming? Yeah, so right now, the way that it works, we started with a fun app that was back in 2020 and just as a job for that, now mostly focusing on connecting people to job entitlements. But of course, quickly work first and then try to connect. So yeah, right now we have a smartphone app, one for jobs, one for housing, which is worth conversational and just to go into a little bit of depth in terms of what a user journey looks like.

Vaissnavi Shukl
So I don’t know too much in detail, but if it’s a laborer who’s using an app, what does that person’s journey look like and if that’s posting openings, whether it’s either for housing, what does that I don’t know, the landlord side of the interface localized. Or do you have a separate interface where you find all the posting work?

Jacob Kohn
Yeah. So as far as housing, we consider building an ecosystem, again, right, building a platform that is helping migrants make better decisions on housing, and we realize both through our respective thesis, I would say that data is quite scarce, especially for informal housing, affordable housing market data we did, Rushil can talk a little bit more about as well. If we engage local champions onboard supply of housing so they go around their neighborhood. These are women, housewives, maybe their husbands working, and what they’ll do is their own community rapidly has been able to do so in the form of a smartphone app. So for them to use this app for photos, details about the property and landlords, then, once they’ve onboarded the supply of housing, we have migrants search for housing. So they’re able to search for the smartphone app recently enabled, please search through a voice note WhatsApp chatbot sends a voice note in their native language, and they can instantly get results from Our database. And that’s you think, is migrants. Often these kinds of solutions are left out of the conversation.

Vaissnavi Shukl
Rushil do you want to talk about the I wouldn’t call it, but the legal policy side of things? What happened to the old housing market now has somebody who’s trying to take away the role of a middleman or a broker, and kind of recheck the system by removing a middleman?

Rushil Palavajjhala
For this particular target audience, there is tremendous trust in a failure human being. They lower literacy in that case, also low. You know, a lot of the users we come across still struggle to navigate like a UPI application. Human beings are critical to this ecosystem. So rather than brokers or, you know, so called “Remote” middlemen, we, we actually were the community led model. We need a human being. Does a human being have to be, you know, an outsider? You know, this reliable unit to women within these communities, these were the wives of these migrant workers. We talked about these industrial workers whose mornings, tending to household, course, tending to the children. They would only find some time. But we found an opportunity here where we were able to train a lot of these women on how to use a smartphone and then go around to find rooms and properties.

Rushil Palavajjhala
Picked up was they had to trust they already had their socializing circles, meet each other and exchange information. In fact, there’s a paper by J pal right which talks about gossip, the term gossip, and literally classifies it as some sort of data in communications. Communication, you’re archiving data to retrieve it when you need it. So in some sense, we literally went through those kinds of channels, and it created great employment opportunities for the women, because they were being paid, you know, per task, like they onboard a house and they take pictures, videos, etc. So some spectacular success stories where there was one woman who was, you know, and, yeah, last I heard, she was making an average of 25,000 rupees a month, quite a bit more than a husband as well. And which was paradoxical, because in the first community, this is a community of migrants and narrow London, they were prohibited from leaving the neighborhood without and the moment they saw money and gain, you know, they were able to use this economic mobility to kind of tied over a lot of the social restrictions and the community and the men, particularly working and their skill sets, etc. So I think it was a great story in that sense. And now coming back to some of those so of course, brokers, I mean, was it really a challenge? Because the community kind of filled that role, and the community is serving when it comes to, I don’t know if you’re talking about the semi formality of the housing, when I say the state, I say the whole ecosystem, you know, which includes the low level bureaucrats, the middle bureaucrats, the politicians, everyone. Governments have a term called regularized. Simply means, yeah, we know everyone’s done something slightly illegal, but you can pay a fee and call it clean, right? Because they like, oh, that’s a revenue opportunity for us. And this is the case, right? The way land typically becomes urban is, you know, there might be an agri piece of land that might have been sold by a farmer, probably might have been subdivided into plots, but you go on using probably 56 subplot divisions, and they have some sort of agreement in value of stamp paper, etc. It’s not entirely illegal. There is some way to track there’s some way of paperwork. But these are, these are shady women who say, Okay, let’s recognize it. Let’s start. Let’s at least give them an electricity connection so they will make the municipality recognize that. Yeah, we layer. And then eventually, you know, at some point, the government decides to lay a pipeline, and that’s when they decide, oh, now we are fully legal, so we also kind of play along with that. Ask people for information only that they’re comfortable with, don’t worry about information, wait for a migrant, you know, potential tenant, to find that opportunity and be able to kind of occupy it, right? So that’s that and a lot of these are facilitated by these women from the community.

Vaissnavi Shukl
So it seems like the agents at play here are right, right from your migrant worker to people who are posting either jobs or housing vacancies. Then on your control, right? So, of course, the job market is not in your control, nor is the housing market workload to connect to market. What are some of the obstacles that you faced when you piloted one do, whether it’s relating to a particular area or a particular state in India, and how, whether the job market or the housing market, or even at a policy level, things that were beyond your control have somehow come into play In terms of how you are working towards your goal.

Rushil Palavajjhala
Jacob can talk a bit more about maybe we can start with the tech side, especially our use of AI, some of those gaps.

Jacob Kohn
So I think you know, what also drove me to India for the first time was just this acknowledgement that India is so data rich with some of the opportunity to connect people different, without reliable data, you can solve a problem that we face in terms of getting difficult to take something that’s informal, unregulated, non standardized and turned into something that’s actually but as a data scientist, and so we get used to understanding how to attack a problem from multiple sides, to get some kind of answer that will help you answer your research question in terms of obstacles. So we thought, you know, going into having a campaign with posters, started out putting QR codes on these posters. After about a month or two, we realized nobody scanning these QR codes on the community and like people’s recognition that a QR code means UPI that it means action that you hadn’t really thought about this, even down to the village, neighborhood level, people understand what these QR codes can be typically used for. So that was interesting in terms of, I guess, the larger picture of technology access, I think, really just just AI being able to solve these very difficult kinds of human agency problems. So Satya Nadella, the Microsoft CEO, recently talked about it in February of this year, briefly mentioned us talking about how, you know, there are all these use cases in India that are trying to solve through AI. I think it’s a tremendous opportunity for us. Language can instantly connect to the services that we’re offering. So it’s about meeting users, where they are, understanding what kind of technology barriers, literacy barriers, are their experience, and trying to meet them as close to that as possible. Mentioned, it’s not just a technological approach. You have to actually be on the ground with understanding their concerns, also just being friendly, so that any questions they’re not just going to download. You have to build that kind of trust technology.

Rushil Palavajjhala
But I can probably dive into the market side of so called, I mean obstacles per se, one rigorous in you know our data analytics look like.

Rushil Palavajjhala
A few months in, like, one and a half time wage rate, you know, what was it, did I say that money go back home, the village. So, they were not going to be the housing market, they were going to be someone who goes through the same. The job was at the same location every day, and typically in England, that was the industrial sector so again, we started looking at industrial arts and housing stock available, and the transport network. So of course, there was no close in terms of scoping communities, we by creating these women entrepreneurs, in some sense, we created my out of there because we got them digitally literate financial and they have their own intelligence and their own networks, so the moment they know that they want to run out of housing on board or manage their area. To start looking at new areas. So in a way, some woman was from Pune, and said, hey, you know what is a great opportunity sitting at home or in your neighborhood? So training related by us, but over zoom calls and some one of these people also, she migrated for a second or third time, like her husband Delhi. So she migrated. And then, since she’d been working on this product for one year, you know, implementing it, she decided to start implementing it in Delhi. Also in Delhi, for example, we didn’t have any staff to begin with, but it literally went viral, in some sense, just based on so it’s grown quickly on its own, very organically. We’ve got almost, we got zero marketing spend, and that’s been working well. But of course, we also faced obstacles in the areas. So let’s say, like an open secret in Ahmedabad, for example, that the largest housing inventor by fuel landowning, traditionally have access to pasture land, and they’ve also constructed housing that they wanted to so that this is a great opportunity to aggregate kinds of housing in one go with the with one visualizing or, you know, I mean, for a variety of reasons, compliance and regulatory. Is it any attempt we had to make indoors in those areas that’s all quite literally violently met with. If you would put up posters with QR codes, I mean, forget scanning to tear those posters off. Within an hour, they would disappear. Some members were quite insistent. We got threatened with lathis and all that. Right, so that’s, that’s where it got to. So we said, you know, let’s just leave these areas alone. There is tons of other inventory. And so that’s how we’ve gone, even in terms of approaching governments, you know, we’ve come very close. We, of course, have done a couple of government partnerships, very close to an agreement with a city government, you know, soil faulting, get stuck, you know, with the cycles of standard committees and the success and changing government. We were, you know, close, I think, with the politics and priorities, it keeps changing. You know, there were some very petty things, like there was, for example, one state in India is one of, like Kerala, for example, one of our most thinking in terms of, you know, manual migration to the Middle East, etc, is also higher. Moving from, say, Odisha etc, to Kerala. It is broadly encouraging the government’s, you know, they’re also, at one point talking about migrant hostels etc. It just took one or two in migrant worker time. And, you know, maybe suddenly the entire medical narrative became anti migrant. So then it was just a thing to do, and just take a small change. We were trying something in another southern Indian state, only there were floods, and the entire state machinery became, you know, focused on mitigating that disaster. Migrants don’t vote, right? So there’s no political brownie points we made in the destination area; they vote in the source areas, which generally are not the ones who are servicing them when they migrate. So of course, these are the kinds of, you know, very systemic challenges on who is interested in solving for associations like the CI example, example, took up this aspect and said, Let’s do something for them. It, of course, also came from business that they needed workers. But the moment labor supply situation reversed after the pandemic again, you know, it is not top priority for a lot of people, so I think that’s what you know, we are still trying to find a major force that would push things in this ecosystem. And one of the more established models is to level the line, and that’s also where they’re focused. There’s tons of housing, urban housing in India is informal. Nobody knows about the cash flows, nobody knows about the asset quality. Understand they are actually going to become formal, or they at least hold formal jobs. So can we leverage that data, rental cash flow data, and build institutions? So yeah, you’ll find motivations of capital and then try and find a way for them to reach what you think is important, because maybe you’re the only one or you’re really stupid, or maybe the only people interested in solving it. It’s interesting.

Vaissnavi Shukl
Something that another guest, eta mentions, and heard what is, how do you really work around these affordable rent rental housing create a part towards ownership. So very serious. Do you really build a lovely set from one person to a husband and a wife and a kid so they’re able to move but also eventually claim ownership where they’re living. And of course, because employment and migration is cyclical, I’m sure you might be back with the platform that being said, what next in the pipeline for guys? See it growing? Or what other areas do you see in one direction to the business side?

Jacob Kohn
Of course, the opportunity is to leverage this. You know, land in India is probably not going to be in lifetimes, or at least in our activity. And because supply of land, land is a limited resource. Conversion of land from one to another degree versus urban. Keep talking about jumping into it also was that democratic interventions, smarter cities or infrastructure, talking about migrating urban, urban. I think getting into that is exactly where we’ve been focused. So I think more and more If India is urbanising very rapidly, all of us are focused on the supply side – technocratic interventions… no one is really talking about who is migrating, who is urbanising?

Rushil Palavajjhala
Is moving this model, and you know, a pathway to profitability, especially for a startup, the pieces you’re most excited about, especially building up some of the more cutting edge products, data and behavior. We can also see a lot of shift again, from the gig economy sectors, several other sectors that have traditionally engaged with this target group, but not necessarily intervene very directly in solving, you know, it’s become gradually right from having no insurance, fine, you know, insurance and trying to see how some other lifelong can start coming in, or the technology piece, Of course, I think they’re super excited. We’ve got a, yeah, maybe Jacob can expand more on that.

Jacob Kohn
I mean, a lot of us are following just to develop digital public options, but much more recently, in terms of AI models. Large language models are very rapidly leveling the playing field for many people in India, just in terms of communication. So as these language models continue to develop, we can’t see a lot more languages coming online to serve India’s very diverse population and Bundu, you know, we’re trying to get cut to our people at the local level. Just kind of use AI, which, you know, is a popular shiny object, really thinking about it critically as where it makes sense to plug in as a key use is to actually empower people who traditionally. So using it to solve very tough problems with transactions. So I think it’s very time.

Vaissnavi Shukl
I mean, especially I read,Shoshana Zuboff, who teaches at HBS, about surveillance and capitalism, not saying we all use data for the wrong things. But it seems like there is a lot of data that is yet to just be formalized and to some extent, to see how you guys venture into new territories and how you kind of keep adding different models to it. And look at this whole demographic that it’s probably not so serve holistically, through, through other means, as you guys are trying to do. So thank you guys for your time and wishing you all the best for Bandhu.

Vaissnavi Shukl
Special thanks to Ayushi Thakur for the research and design support, and Kahaan Shah for the background score. For guests and topic suggestions, you can get in touch with us through instagram or our website through our website archoffcentre.com, both of which are ‘archoffcentre’. And thank you for listening.