Listen to the podcast to learn about

1:35 - Putting the entire developing world onto Google Maps

4:00 - How the regular public helped build Google maps

9:00 - What is account aggregator? What is its objective? What is the motivation behind account aggregator? It’s use cases

19:10 - Business model & opportunities for account aggregators

24:16 - Things account aggregator will change in the future

28:10 - What attracted Sanjay to work on Aadhaar and what were his expectations from it

Read the full transcript below

Shripati Acharya 0:22

Welcome to Prime Ventures Partners Podcast. I'm Shripati and our guest today is Sanjay Jain. I'm absolutely delighted to have Sanjay with us. He was the Chief Product officer at UIDAI. Before that, he ran a product at Google Map Maker, a crowd sourced mapping product that helped to build Google Maps data for over 170 countries. He has been among the key people who helped develop India stack, an API based infrastructure that allows technologies to be built around Aadhaar authentication, E-KYC, E-Sign, Digital locker and of course UPI. He's also a core volunteer with iSPIRT, the software product industry Think Tank. And now, he's a partner at Bharat Fund, which invests in tech to drive inclusion. So, delighted to have you with us. Welcome Sanjay.

Sanjay Jain 1:16

Pleasure to be here Shripati.

Shripati Acharya 1:18

So I'm going to start off with your journey at Google Map Maker. So tell us a little bit about what was the problem, which you're trying to solve at that time, and also around what time was this?

Sanjay Jain 1:33

So I joined Google in about 2004. And at that time, Google had just acquired this small company, which eventually created this product, which became Google Maps.

Shripati Acharya 1:44

Okay.

Sanjay Jain 1:45

And so when they launched Google Maps, it only had the US, UK and the rest of the world was blank. And they were doing it on the basis of licensed data. And so we said what would it take for us to get this pretty cool new product to support us in India. And we tried to find license data and we were unable to find it. Meanwhile, they're also acquired this other company which had satellite imagery from around the world. And we thought it'd be a great thing if we could put the two together and people could look at satellite imagery and draw maps. Because we just couldn't get good quality data. We said, Why don't we let users map their own neighborhoods? I don't know if you remember back in the day, if you wanted to send out an invite to people to come home for, say, a birthday party for kids or something. You always had this little image with a hand drawn direction saying you come here, you turn right turn left, or whatever. And if you could do it once and for all on Google Maps, then you know your problem was solved. So we thought people would do it for their work, for the homes, for places they visit often. And that would be sufficient to get us usable maps for the country. And so that's what we worked on. We built this product, launched it initially for India and then eventually for developing countries one by one and eventually we put the entire developing world onto Google Maps through this product. So Google Maps contain licensed data for the developed world and Google Map Maker data for the rest of the world.

Shripati Acharya 2:04

Wow! So, I think it is one sense the whole concept of having digital maps, probably in most of the world came first through Google Maps.

Sanjay Jain 3:21

Yes. And this was how we got it for pretty much everyone except the developed world.

Shripati Acharya 3:26

So how do you get, while the concept is good. I can just imagine you sitting across the table and discussing this, that Okay, let's actually crowdsource this issue. How did you get about getting the word out and the contribution from people, you could even do something like this

Sanjay Jain 3:43

Yeah, I mean, we’re a fairly small team. Lalitesh Katragadda and I did most of the work on this. And we were clear that once people knew it existed, and they would then do it and they will tell their friends and because that would help improve the success rate. And in initially some of the management believed that we might need contractors to help build it out as well. So we had a mix of both. So we actually did some marketing actually used Google AdWords, got the word out about Google Map Maker in some parts of the world. And we spoke about it and I think a bit of organic work helped. And also I think the contractors helped as well, a little bit. So we launched some countries with some basic data filled in so like, you will get the national highways in the cities marked, the others are completely blank. And people filled it in, everything from cities, national highways, rivers, railway lines, everything.

Shripati Acharya 4:41

So how did you actually make sure the data is correct, because I'm not a cartographer, I actually am enthusiastic about it. So I just go and map something.

Sanjay Jain 4:49

So as long as you map something you know and it's true, because you're drawing it on top of satellite imagery, so as long as your intentions are right, I think you will get it right. The second part is that nothing went live until multiple people saw it and approved it. So you could say, I know this region well, and I'm going to look at edits in my region and approve or disapprove them. So you could do quality control. We had some of our contractor team check for spam kinds of things. Yeah, of course, you know, people do all sorts of things. And then we built in a lot of machine learning type algorithms to detect what was potentially a good edit or a bad edit. And so all of that was built into the product. And that's how you got the quality up.

Shripati Acharya 5:36

So you must have had some crazy stories in the initial days with respect to the users who went around mapping their neighborhoods?

Sanjay Jain 5:44

Yeah, I mean, we had some amazing people. And we actually got to meet some of them. At some point we got all the top contributors into one place and we had an event with them, and you got to hear the stories and it was fantastic. So we had this retired civil servant in Kerala, who mapped a large part of the areas he knew. And then there were other people who were mapping everything they could lay their hand on. And it was fascinating to watch. If you sort of start from an empty slate, somebody marks a road, then suddenly people mark the roads to the side of it, they mark landmarks on it, and so on. And it starts to become dense, people just add more to it starts to acquire the right character. And we, in fact, at that time, set up this mechanism by which we recorded the data as it went live. We built time lapse movies of cities going live. And we've put a couple of those out on YouTube and they're fascinating to watch. You just see an entire city come up on a blank map. We actually put them out as YouTube videos after some time, and it was just I think, fascinating to watch how countries came onto the map, people started to use it. Initially, if we're not sure the data will be good enough for driving directions. So Google actually delayed driving direction launch on user generated content for a very long time. And only when the quality got where they thought it would be okay, they launched driving directions. And today, everybody is using either Uber, Ola or whatever is getting driving directions from this data.

Shripati Acharya 7:28

Yeah, in one sense, many of us might not have even believed that something like the maps in India could ever be digitized, because they're not like nicely planned grids as you have in the Western countries. They're all curvy squiggly lines in terms of our roads, and just getting the driving times right. This is amazing?

Sanjay Jain 7:49

One thing about India, you discover a lot of roads without names. It's not common in the rest of the world. At least in the developed world, it's very uncommon to find roads without names. Other than the hurdle to driving directions,

Shripati Acharya 8:02

Yes! So how did you deal with that?

Sanjay Jain 8:05

So this happened after my time. I had already left by then. But, driving directions are of the form that you go x kilometres and turn right onto an unnamed road. Or the landmarks then became the next step. So actually first that was how it was done. You just got like 250 meters right turn at an unnamed road. And then it passed the petrol pump or there's some shop in the night or whatever. So, you actually got directions with landmarks at that point, because that was really the way it turned out to be.

Shripati Acharya 8:37

So let me fast forward the journey a bit, to your time at Aadhaar, which is really the foundation for all the work on top of our India stack. So I really wanted to come all the way to the current time, about the account aggregator, I wanted to really spend a good amount of time on that. So tell us the motivation behind account aggregator, what is it and what is the objective behind it?

Sanjay Jain 8:48

So if you see all of us do various transactions with different service providers, and typically data about these transactions resides with that service provider.

Shripati Acharya 9:12

Share some examples of that.

Sanjay Jain 9:14

So, for an account aggregator a good example is a bank account. I deposit money and withdraw money, I write a cheque, I make a payment, all of that information sits within the bank account. These are transactions made by me which sit in the bank. Another example would be telephone calls. Telecom company knows what calls I'm making, they give me a monthly statement, but the data sets with them. Now, this data is quite useful and valuable, both to me and to other people who might want to provide services to me as well. And what an account aggregator does is, in some sense, it's saying that you as a user has rights to this data, and that the service providers should be able to provide it to you in a format that is usable by you and by others. So essentially, what the account aggregated does is it provides a mechanism for that data to be brought to you under your control. So every service provider in the financial space, the regulated space is required to now provide this data to you. And that's what account aggregator does, it provides a mechanism to do that.

Shripati Acharya 10:21

So can you give an example of what's the use case for this, which was envisioned here.

Sanjay Jain 10:26

A few use cases, a simple one is lending. Today, When I go to a bank to apply for a loan, the bank asks me for data bank statements. And I typically go to these banks they print out and give it to them. They go ahead and digitize it again. And then make some sense of that data. Now, what could happen is that with this new infrastructure that's created, I could go there and tell all my banks that, hey, this is a bank that I want to send my account statements to, could you please make that happen? Okay, and then give my consent to the account aggregator, the account aggregated, takes it to all the banks accounts, and makes sure that the data gets transmitted to the bank that I want to take a loan from. And that's essentially what the account aggregator does is it takes my consent and goes to the service providers. And then it gives me a report saying the data was obtained from these guys at this time. So it tracks consent flows and data flows, the actual data doesn't flow through the account aggregator. Only the consent mechanism does.

Shripati Acharya 10:42

Very interesting! It's like the control plane network, control plane and data plane.

Sanjay Jain 11:55

Exactly! So that's all, so this is just a control plane, and that allows the routing of data to happen. And the use case would be personal finance management. So I could give all my assets history and so on to some software, it would look at it and say, here's the right investment strategy for you. And then you could execute that strategy. And then it could monitor and assist you to make changes as required. So that could be another example of this use case where again, in the finance industry, is a very common use case. There are other places where you use this data, I go apply for a visa, I give them bank statements to assure what is the balance that I have in account & so on.

Shripati Acharya 12:37

In the visa case, just to clarify, the Embassy of whichever country you're trying to go to work would make a request, you'll get a notification saying, hey, do you approve this request? It goes to SBI or HDFC or whatever your bank is and then it goes back to the embassy.

Sanjay Jain 12:52

So that's sort of the logical way to do it though the way it is set up right now only financial information is regulated. So what would happen in this case is that as a user, I would say, hey, I want the data. To show up with me, and then I will hand it to somebody because embassies are not regulated entities.

Shripati Acharya 13:13

Got it! So, they can't actually access it. They can't ask the account aggregator to get them data. So what about the format for this data? Are they all like the format and the standards here been published and agreed upon by all their different constituents?

Sanjay Jain 13:27

So it's not fully done yet. But the way it's set up is account aggregator itself is neutral to the format since it actually doesn't even see the data. It's kind of agnostic to that. What we're expecting is that all of these players will realize that the banks who are providing the statements are the same banks who are reading statements. So, they will eventually get to where they will standardize this. Okay, but step one is actually to just make the transactions happen and to make sure the user stays in control.

Shripati Acharya 13:59

So in one sense, the contract aggregate doesn't mandate a particular data format per se. What it mandates is that you need to have these content flows. And then after that between the user and the provider, they can have their own format.

Sanjay Jain 14:14

Well, chances are the provider, provides some format and the entity that's reading it will read it.

Shripati Acharya 14:20

Fair enough! So what is the state of the union of that? When will we see use cases which are another there is sandbox and so on and so forth? And perhaps you can just give us an update on where it is with respect to the rollout?

Sanjay Jain 14:31

Sure. So when this was being designed, the RBI was also coming at it from a slightly different point of view, when they were looking at it, banks essentially went to them and said, Hey, we have this issue today for this data sharing we're asking users to give their usernames and passwords for net banking. So when we communicate to the user, when we are one way we communicate, hey, please don't share usernames and passwords. Then when we are a lender, we've tell them the exact opposite. And so we have this Jekyll and Hyde view that gets users very confused. And we don't like it. So can you help us with it? And so they were going down a certain path of how to do it. And then we felt the right way to do it was with API's. And that's actually how this concept was born. And they then created this intermediary who's now fiduciary for the user called the account aggregator who makes this happen. So, that's the concept that came about. So when they did this, they asked for people to become licensed entities. Many applied, seven have been approved. And so, the way it works is you get an in principle approval, then you go and build your systems, you get them tested, get them certified for safety, security, whatever else RBI requires you to do. You submit compliance and audit reports. And then RBI approves the system to go live. Of the seven entities one has gotten a formal go ahead to go live. Now, but when this goes live, you want the bank ecosystem to be ready as well. And so these seven entities have gotten together, they've created a group called "Sahamati", which then has been working with the entire ecosystem to make sure banks are ready to give the data as required and so on. So they've been working with the large data providers as well. And many of them are ready to go live. So I think the plan is within the next couple of months, some of these large institutions will start to go live with it. So it will take time for the entire ecosystem to go live for every bank to be on it for every and it's not just banks, it is every institution which is regulated by RBI, SEBI, PFRDA and IRDA. So insurance, mutual funds, pensions, and banks.

Shripati Acharya 16:49

I had heard some things around this also being extended to health care, education, for all those things which are not regulated by these organizations.

Sanjay Jain 17:00

So, the account aggregator is an RBI construct, because the regulatory bodies have the financial sector regulators all have this council which sits over the seven financial sector regulators and they got together to approve this. So non financial sector entities are not part of this. However, if you look at what the data protection bill contains, it essentially has the same construct in the law. And it is expected that other regulators will either create their own or some administrative changes will allow entities to become aggregators in those domains. And then this entire ecosystem will go live with other data as well. So, you can think of telecom data, you can think of healthcare, because all of these are regulated sectors, education, all of these can start to come in.

Shripati Acharya 17:50

In one sense, I would say that if you look at the New Age lending, they would actually like to use this data as part of their credit scoring, isn't it like telecom data, education data, all these things?

Sanjay Jain 18:01

Sure! They would, but in terms of regulatory bodies that's going to take time to happen, so each regulator have to act on this, actually once the Data Protection Authority comes up that makes life easy, because then while RBI took this step as part of their own decision making process, the bill essentially will mandate people to allow people to access their own data. And when that is done, then this mechanism will hopefully get adopted everywhere.

Shripati Acharya 18:32

Okay! So just in terms of some time frames, just to put some dates out there. What would you expect when we can start seeing somebody who is enthusiastic about this as a user to actually be able to use this to get any financial products loans, for example?

Sanjay Jain 18:49

Yeah, so financial products, like I said, in the next two, three months, the launches will start so the account aggregators will do their launch, the banks will keep on adding into the mix. So and then the other providers start to come in as well. Okay, so I do expect the long tail of providers to take time to come in. So for example, the major banks, I think, are all mostly agreed to be a part of the system. So they will come on board very soon, but you can expect the smaller banks to follow a little later. So we're talking about a few months for the first system to come up.

Shripati Acharya 19:24

Okay! So let's talk about, I mean, this question has come up a number of times whenever account aggregator is discussed, which is what is the business model for the account aggregators, and the reason this question comes up, of course, is that if there is not a viable business model, then why would the private account aggregators provide the service? So what is your sense, of course, remains to be seen exactly how it plays out, but any thoughts on how that might evolve.

Sanjay Jain 19:50

So, the first use case is lending. And in the lending use case, what we've seen so far is that the lenders are willing to pay for this data today. They charge a processing fee actually to the user as a part of the loan process, but then part of it is used to pay for the data. So, if you sort of apply for a home loan, you pay a processing fees, and part of that actually then goes to the people who are getting data for your loans. So in some senses, users are paying for it but not directly.

Shripati Acharya 20:18

So bank is actually going to these data providers and saying, hey, look we want this data, and here's the fees for that.

Sanjay Jain 20:25

But sometimes it can go back to the user and charge him with processing fees. I'm sure many of the small costs are covered in that. So it's not just that fees. So in so many ways, the user is still paying for the data. But the point is that different use cases will find other parties willing to pay, but in some use cases, you may not have anybody who's willing to pay. So this is going to take some time to evolve. Because, even today in the lending case, users pay for their own data, they don't realize it, they just see it as part of a processing fee, versus being suddenly told by an account aggregator that hey, this is going to be 500 rupees or 30 rupees or whatever, they don't realize it, but it goes as part of a 1% processing fees that you're paying the bank for the loan.

Shripati Acharya 21:09

But maybe you still continue that way. If I'm actually getting a loan, I could now make it a half percent processing fee, but because it could be dramatically lower, maybe even less. And then I pay on your behalf to the account aggregator. I think the user is paying, where it comes from is different.

Sanjay Jain 21:26

So there are two or three elements to it which are going to be important, one is to remove friction in the process. Because if the user is suddenly confronted with a fees that is unexpected, they are not used to paying, that might be a friction. So, everybody will want to try and avoid that friction. The second is the fiduciary responsibility remains of the user. So even if somebody else is paying for the data, in actuality, you want to make sure that, the service provider knows that the real customer is the user and not that other entity. Because that’s the very reason why a separate entity was created. Otherwise you could have had a bank talking to a bank and getting the data. So the reason why you put an entity in between was to say that this entity has fiduciary duties towards the user. So, you want to preserve that model. So but yeah, the money flow might, to reduce friction might go in a different way. But in terms of AA, the users are always going to be the customer.

Shripati Acharya 22:23

Makes sense. So, what you’re saying is that for a loan, the user might pay a small fee to their account aggregator, and then that enables the loan to be processed faster, quicker. And maybe even cheaper, any other way we could actually see this evolve.

Sanjay Jain 22:37

So, there are three players in the scheme. One is the entity which holds the account. There’s the entity which wants to use the data and the user. There’s really three entities. And you could say that, you’re a preferred customer, I’m willing to think of it in terms of the basket of services to provide you. You have infinite access to your own data. I could do it as a service provider. I could do it as somebody is providing a service to the customer saying, hey, I’m giving you the service. And obviously, this is a commercial service and I’m making money off of it. I’m willing to pay for this data access, because it gives me a lot of comfort to have accurate data in a timely manner. And that reduces my costs, etc. So I’m willing to take some of the benefits I get and pass them on in this manner. And then the third is where the user says, Hey, I want to monitor my own account. This is a simple way to do it. And maybe I’ll pay a small subscription fees and get access to my own data.

Shripati Acharya 23:37

Makes sense, or maybe it will be a combination where in a provider certain conditions.

Sanjay Jain 23:43

Like for example, you’ve seen this with credit reports. So normally, what happens is the lender pays the credit information company a fee for the report, but the users can get it free once a year. So you already have this combination working out there where, in some cases, third parties, paying for it. And a user gets it once free and can get it multiple times at a price.

Shripati Acharya 24:05

Exactly. It is also like ATM withdrawals. Like you can get an X number for free. And after that you got to pay for it.

Sanjay Jain 24:11

I do expect a hybrid like that to show up because it makes sense to do this.

Shripati Acharya 24:17

Make sense. So essentially, in the next two, three months, by the time summer rolls around, we should start seeing our greater use of it. So what will get most impacted if you see like this looks like it’s changing the way business is done in financial services, of course, where loans are a huge part of that entire landscape. So what do you see as we’re going to get disrupted here, as AA becomes more mainstream?

Sanjay Jain 24:39

So first is that expect lenders to get more credible data. Right now they might get data which comes through a user they’ve spent a lot of effort to check it, double check it, this data entry, so all those pieces in between will go away because now you’re getting data from the source.

Shripati Acharya 24:58

Makes sense, in a digital format.

Sanjay Jain 24:59

In a digital format, so you’ve already taken out significant costs from that. The second thing that would happen is that some of this data will become very useful for monitoring. Today, you don’t monitor accounts in the same way. So I give loans to somebody, but then I don’t necessarily, because it is such a pain, I don’t have the capability to monitor how the business is doing on a regular basis. So, I could actually create a cheaper loan for somebody and say, hey, I want the ability to monitor your credit on a constant basis. So every month I’m going to pull a credit report for you based on all the data you have. And if your credit starts to go worse, maybe you get a friendly call from me saying, hey, okay, so you know, that kind of relationship. So far, you haven’t ever had to do this, but suddenly that is becoming the new norm.

Shripati Acharya 25:52

New ways of monitoring the credit.

Sanjay Jain 25:53

Correct! Will come into being and so that would be one thing, the second thing that might happen Is somebody might say, I’m gonna build a model for this model, I need data, I’ll pay a million users some amount, qualify them in a certain way, and then build a model based on their data, then use it for something else.

Shripati Acharya 26:15

So, in one sense, people are free to monetize their own data.

Sanjay Jain 26:21

So, I think we will start to see interesting things happen, because when you start to free up this data, you will see that people will find innovative uses of the data and be willing to pay and all be clean aboveboard, the user will always be in control in the path. And which is perfectly fine.

Shripati Acharya 26:39

It is very interesting. Put it another way that, the centralized repositories of data, will kind of become more porous, because now the user will have the ability to take data from all these repositories and monetize it, because ultimately, it belongs to them and do what they wish with it.

Sanjay Jain 26:55

And sometimes it could just be better service. So, for example, Google personalizes my search history because they have search history. So for me to switch to a different service provider is difficult because I’ll start off with a crappy search quality. But if they had my search history, maybe they could have done a better job doing searches for me. So logically, I could take this. And Google does allow this, by the way, so I don’t want to say it doesn’t. But logically, I could download my search history from Google, give it to this new entity and get better service from them. And so the same concept could now start to come through in terms of data is that, if you understand me better, maybe you can give me better service.

Shripati Acharya 27:39

Makes sense kind of reduces the barrier for new entrants to come in and start providing our services. So we’ve come to close the podcast, I’ll ask you a couple of other questions, which is, what is the one thing which you have seen this entire platform evolve over a long period of time, from the from the days of Aadhaar, so what is it that you thought was going to happen? And very sure about seven-eight years ago when this entire process started, maybe even longer, which isn’t true anymore in your opinion, or which has been radically transformed?

Sanjay Jain 28:17

So when we started out on this journey, I mean, part of the reason that I was attracted to Aadhaar was, you’ve seen what’s happened in the rest of the world. ID has always been a basis for development. I mean, you know that the financial system in the US works because all that data is linked together to the social security number, and that allows for better fraud checks, and allows for easy credit. Many things become viable because there’s an ID in there which links this data up. And, whether the user has control consent or not, has historically evolved in different path than we have evolved, but that essentially makes life easy for many people.

Shripati Acharya 28:59

Like Primary Key.

Sanjay Jain 29:01

Correct! So I did expect to see us go along that path and see a lot of formalization happen. And with this thing being the way by which data came together and made people’s lives easier, and some of that has happened, but it’s not happened yet in the same way as it has in the US. I think it’s taking us longer. I think the privacy conversations have changed the way people think about it. But at the same time, it’s evolved us. So I wouldn’t have thought of the account aggregators being an element when we were looking at building Aadhaar. But today it is. I could have imagined the Digilocker but not account aggregator. Where I have a document tag with my ID and my ability to get it. And but not this process of me being able to assert ownership of my data. So I think we’ve been positively surprised by that. In terms of usage I thought Aadhaar usage would evolve and happen much faster than it did. So while the government sector, of course, took it up very quickly, the private sector was slow on the uptake, which was a bit of a surprise. I mean, Jio of course, the telecoms and they did use it very well, but the small startups, etc. That’s actually why we created the fund. Well, some of it was to see why, when you have these new platforms, you’re not seeing as many startups building on these new platforms. And perhaps you thought there’s a lack of risk capital. But that was a bit of a surprise to us, because you think of tech geeks and as these risk takers who jump on anything new and use it, and it didn’t happen. So I think that was the bit of a surprise that we don’t have more startups today who are doing it, this is a conclusion I’m giving from before the Supreme Court ruling Aadhar. And post the Supreme Court ruling on Aadhaar, of course, this would not have been possible. But you’d have expected that before that happened this would have.

Shripati Acharya 31:12

Which is essentially paraphrasing it using Aadhaar as a primary key for storing all your data. And all that stuff.

Sanjay Jain 31:17

No we don’t want to use the word primary key, but using Aadhaar to onboard customers for lower cost. Even back then, our fundamental advice to people was to use it to onboard people, but then to have an internal primary key, which is separate. So for example, if I open a bank account, there is a bank account number, some kind of account number. There is a customer ID in the system, which is within the CRM. And one of the attributes is ID. As opposed to the ID in the primary key. You want to make sure that your CRM entry has its own key, and you can deduplicate it on the ID as well. But that’s not the primary key. So conceptually that’s what we had always encouraged even back then when we were thinking about how you would expect it to be used, which is to say, never keep attributes as primary keys. As a design principle, that was, the other thing that you’re seeing today. And it’s not taken off yet, but conceptually it will, is more decentralized systems.So you even today, if you think of other the decentralized ID systems around the world, they keep user data within control of the user, but the data they have includes IDs issued by the government. So the government system, of course, is centralized. But then I have access to my data store, which then allows me to use it in different ways. And that is a concept which I think with the account aggregator, giving us one way to get access to your own data, Digi locker having data tagged to you coming in a different form. I think it’s all an evolutionary path. But eventually, I would like to make sure that as a user, I have full control of all the data that belongs to me. And you know, we are on this journey were, will be multiple things to both enable it, and to push it forward. And so I think that the stage.

Shripati Acharya 31:56

Well, thanks a lot Sanjay for your time and delighted to have this conversation with you.

Sanjay Jain 33:24

Thanks! Shripati, this is wonderful.

Enjoyed the podcast? Please consider leaving a review on Apple Podcasts/iTunes.

Follow Prime Venture Partners:

Twitter: https://twitter.com/Primevp_in
LinkedIn: https://www.linkedin.com/company/primevp/

Let us know what you
think about this episode