00:00.54
Narayan
Hi and welcome to yet another episode of the new space india podcast today if he we have here with us gunjin who leads Thya Earth Gujin welcome to the show and thank you so much for taking the time for speaking with me.

00:13.91
Gunjan
My pleasure Naran and really glad to be here and speaking with you and looking to learn from you from your experience. Also.

00:22.95
Narayan
Absolutely I mean you're very kind incing that. But I'm here to learn from you. Um now ah I looked at you know what? you guys did at Tha a few months ago I guess and it was very interesting because I always try to look at.

00:34.59
Gunjan
Um, yeah.

00:40.42
Narayan
What kind of innovation is happening in India and what kind of teams are building what kind of products and you instruct me with the kind of work that you were doing which was very very interesting and I thought I should connect with you guys and I reached out because of that. So from um.

00:46.95
Gunjan
Thank you.

00:52.21
Gunjan
Um, yeah, thank you.

00:56.41
Narayan
You know your perspective I know that you guys come from a non-space background I have stepped into the satellite industry here now. So would love to hear from you. You know how you guys came together with respect to coming from different backgrounds and mechanical engineering and you know, computer science and others and.

00:59.24
Gunjan
Yeah, yeah, yeah, yeah.

01:13.30
Gunjan
Right? right? right.

01:16.00
Narayan
Having done work in the it industry extensively and now stepping into the space industry and what's the story behind Thoughtya coming together as a team.

01:24.80
Gunjan
sure sure soap ah no yeah soap as you rightly pointed out none of us from the traditional space industry. So no, um, ah of a part of us all 3 of us were working in the kind the consumer internet space. Our primary experience our work revolved around Data Science Analytics software development and ah product management. 1 of us is a consultant strategy consultant so he should do a lot of business development work. And digital transformation work for large-scale enterprise clients now. Ah if you go back, try to go back to the story of know how we how Tha started then we'll have to go back to the last organization that I was working in so it was a. Financial research company. Ah online kind of financial research company where we were dealing with a lot of economic data financial data etc and also to monetize our primary model was to know Target our. Ah, users online users with the right kind of products or right kind of ads or the right kind of products so to do that. Basically we had to build a three hundred sixty degree view of our and of our users and one of the variables that we were trying to build in at that time was know ah say of. Ah, proxy for per capita income or proxy for income level of the user so we tried to do that using pinco ah using the ip address David that they were coming in from and we mapped we could map ip at this to a pinco and bit heart from pinco generally from pinco it should be able to gods. Some semblance of what the prosperity level or income level of a user could be so when we step into that domain of understanding ah economic status of a user based on location that is when we were I came across research papers was. Researchers at Harvard were working on to derive poverty level in countries like Africa or sub-saharan areas where data is very poor to understand based on images can we not determine ah poverty levels or income levels at. That's when I got exposed a wow woman if we can determineine such information by just looking at images. Ah what? what other problems can we solve for so that's when the entire research work around what kind of data is available and.

04:12.53
Gunjan
I think it started around 2 years back and that was a time I think set the satellite industry was also moving towards no low low-cost launches a higher number of images being available. So ah, so ah, we saw the anti-industry like anti-eco ecosystem coming forward or coming together and there was a possibility that. Every course have saved speculatmeters of the earth was getting imaged almost every day and ah, ah for us for so all of us comes from so I also have a have a background in strategy consulting I started Mike area at excel expense Saturday so ah, we. Have experience of solving problems for clients. Ah so that's when you've thought you know Bob? what what other problems can we solve for using this dataset so using these images and at money control we had another problem statement of most of the economy data that we used to get used to come with a lot of lag. Um, the data would be of poor quality. Some economies data was not reliable so we thought you know why not can we track economic activities using those images. Ah so ah. We started now economy activities know it's a very ah ah big problem statement to solve for then. Let's start. So we said let's start with 1 domain once a small domain. Let's start with say a stock of economic activities. Um. In the physical space say ah we have three pillars of the economy agriculture energy sector and the metals and mining sector so we believed metals and mining sector was poised to grow. Um based on um, how. Ah, the world was moving away from fossil fills ah as batteries would there more need for batteries there more need for solar panels there more need for metals to so that we saw we thought no metals and mining would be a growing sector so we ah. Ah, we we started our journey with the metalsson mining sector and now using satellite data. We can map the entire supply chain of the metalsson mining sector. We started with Ferris supply chains steel where we can tell you based on images. How much production is happening at any steel mill around the world. At any smelter around the world for copper how much mining is happening or how much inventory of different codes ah or different minerals are kept at different locations around the world. So ah, ah so and how the team so this is how the genesis of the idea know came across.

06:58.25
Gunjan
And we are cofo cofounders basically so myself and niki started this off both of us were colleagues at my previous organization. Ah then I had worked with Naresh ah, who is the city of the company and. Ah, had to I have worked with him in my earlier organization for a very long time and Himman Cho he's the fourth co-founder he is a school friend and I've worked with him for many many years across multiple projectse on multiple projects so that's how the team got together. And now it's a journey journey. It's being around to to to 2 years now close to more than two years now actually yeah

07:42.65
Narayan
Its interesting. Thank you for that Very interesting. You know story of how you guys came together now of course I guess ah a learning curve when you start off into a new field coming from. You know what you were doing with consumer.

07:47.15
Gunjan
Thank you.

07:54.27
Gunjan
Um, yeah, yeah, yeah, yeah.

08:00.23
Narayan
Data and so on to then looking at Satellite Imagery on so when you started discovering that you could start using Satellite Imagery for a lot of these economic indicators and on was it very obvious that you have to use satellite data or you know you thought about using drone data or other such sources to get.

08:07.65
Gunjan
Are.

08:19.31
Narayan
You know, insight into those regions that you were looking at or was it very obvious that satellites were the way to go and if you had a sense of where you would get it how you would get it if you had to pay for it and what was the discovery in that sense that.

08:32.10
Gunjan
Yeah, you're right? So ah, none of us had experienced working with satellite images. We had not seen any satellite images. The only exposure to we had to satellite images was Google Google Google maps ah prior to working on this problem statement. So ah, so. The initial stages were like no ah ah ah now the tradeoff between I think drones and satellite images was very clear to us or very straightforward drones. Do not give lend themselves to the type of ah coverage that we want the frequent coverage. And the extensive coverages that we want so drones was not at all at the picture in the initial stages so it was either using satellite images or other alternative data sources. So our and thesis was in the beginning was alternative predict economic activity using alternative data sources and satellite data was part of that. Alternative data sources. So we started off with alternative data source ah with satellite images and we believe later on we'll keep adding on more data sets like say mobile signals. Um I a if any iot data signals come into the picture. So all the datasets will keep adding but we started our journey. It was clear to us now. Let's try. Because we I had read those papers and it was I believed. No if we can say there was a lot of work that and ah or or um papers written on using satellite images to understand every sector a economy. So I thought you know if some of the physics around. This could also be implemented to the high hit emitting sectors. They should work on the same principles. Basically so we tried a couple of no um and use cases examples kind of how are we getting good signals or not and I mean. Ah, so it's luckily at the first go itself. We got very good results for Legals. So ah, yeah, that's how ah since since we started getting no acceptable results that a first go itself. We then died. Um, we had strong belief. You know. If we further enhance those algorithms um, we should be able to get very good accuracy so we started trying different areas, different plants or different males at different geographies and all of them have to give us a good results. So nope that's ah, that's how it all started then. So. And then it was clear satellite images. So be a way to go then the second problem was nope what type of satellite did from whom to buy what data to rely on so the economy was difficult to ah, arrive at at the beginning now ah say ah for a first product which is measuring.

11:23.89
Gunjan
Production. Ah, we started our journey with Lenssett eight now lens saidt 8 as you know has a repeat coverage of 15 days. So ah, we've tried out our initial products with lenst 8 but obviously um, lens at 8 would not suffice for a commercial great product which can be used by and customers. So and add to it problems of cloud cover one zone so there'll be huge data gap in between. So then we started combining. Landancesett 8 and modi so started usinging lanser and modi so we did some image processing around that we could actually upgrade modi. We could predict lancet eight type images from modi so we started getting data more frequently. Then even though. Ah, fusion is a good approach but ah, it depends a lot on what are the pairs of datasets that we have taken to predict the end output so with lasset eight and mos there'll be huge gaps between the prediction date and the pairs that we take of images. So then we thought you know. Okay, lenset 8 is yeah, did any other option then then we will with end as thermal data at time was not not commercially available with in anyone. My Laance date was the only one I think which was giving good quality thermal data dens. Started exploring sentinel to short wave infrared and we figured out a way around using sentinel 2 swirl bands to basically identify thermal informations from any particular area and then we started switching to sentinel two and sentinel 3 and since. Central to has a cadence of five days it worked well with our environment there. Also we started exploringing know some of the commercial image in a worldview worldview three has a s sower venors I think that's the only ah commercial sower event I think which is available right now. Ah, but I think the cost was very high at that point of time even now the cost is very high so which would not be feasible for us to really expand and scale without ah, no ah and client. So we stuck to sentinel two and 3 and it worked well for us on on. Other products we started off with sentinel to um then we move started exploring other commercial images. Also so it's a mix of both no sent, not success publicly available images and commercial images that we rely on right now.

14:11.60
Gunjan
Ah, so but in the meantime while we are doing this I think it's crazy. But I think of our Usb now we can you know we could work on low and medium resolution image and do not need to entirely depend on commercial images and which allows us to ah ah provide our insights at really at much lower cost. Then what is available in the industry right now.

14:33.78
Narayan
It's interesting again. You know, very insightful. Let's to how a team can build up coming from a completely non-space sector to building your product in the space sector with reance from what you said there are a couple of things that I would love to hear from you as um.

14:41.66
Gunjan
Yes.

14:52.28
Narayan
You know more elaborative kind of a prescription as to how this product evolved now when you look at satellite data and you're deriving economic indicators you want to make sure that it's not like Junkin junk out kind of a scenario.

15:07.48
Gunjan
Yeah, yeah.

15:10.41
Narayan
With respect to data and that people can actually rely on these insights that you are generating and that they are accurate right? So one of the things that you actually mentioned is that they were accurate when you looked at the insights from that. So how do you actually look at.

15:20.80
Gunjan
Um, yeah.

15:27.75
Narayan
Accuracy from a picture of knowing that whatever indicators that you are generating is what is reflecting in the ground or on the ground. So it's the whole idea of ground truthing that you know whatever is being.

15:39.14
Gunjan
Um. Um, yeah.

15:44.93
Narayan
Learned from insights from these machine learning Algorithms or computer vision technology and so on that they are. In fact, you know people can go out in agricultural farms. For example that they can look at a farm and then they can look at.

15:49.41
Gunjan
Um, right.

16:01.84
Narayan
Doing a small crop cutting experiment or something like that to know what is a yield of a particular farm and you know the quality of the images. So but in your case, it's it's something very different in here. So how do you look at that.

16:03.24
Gunjan
Um, right? right? right? right? So ah, in our case obviously ah so of say for the thermal product where we major thermal signals to no measureing. Platform so ah at the onset so the entire platform we first try to build a filter around what type of images can can be consumed by the model. So we put filters around say cloud percentages how many? um what percentage rates of the pixels are covered by cloud. And there. Also it's a big scene say hundred kilometers Hundred Square Kilometre scene we all we don't look at the entire hundred square kilometter scratch stretch. We just look at our Ai areas and see you know? Ah, what are the areas or there are pixels or what percentage of my Ai is covered by cloud so that that those are the basic filters that we put in when using our data. So now. Ah as we had monitored say hundreds of assets already and for those assets we have also from the ground created a lot of ground road information from our partners so that we know the actual productions actual productions. Both reported and actual actually so their numbers which are reported by the agencies by the companies and the actual numbers could have some deance from them which some of our partners know about so we rely on we relied on them at the initial stages to. Fine tunena algorithms or to build our algorithms and now that we know for each unit for each asset. We know at what range those numbers would like all right? So when we produce a new data series. It goes through rigorous checks. Um, okay, ah what percentage of deviation that we see. In that new data series from the ah previous say four five months or last year's at this point of time. So all those checks that we put in while um, taking out any new data series and giving out to our clients on the inventory site or where no where and we do object detection. It. It's pretty clear because we can also look at the images manually ah on top of the checks that I already mentioned we also look at the images manually toin know whether the predicted masks whether the predicted results are in line what we can visually see also if we see a larger larger variation in our. Ah, give an output so it's a mix of no ah mathematics some propriety data sets that we have been able to gather and some manual interventions in terms of looking at the end output looking at the images that we know and shock it. Whatever data.

18:56.60
Gunjan
It goes out to a clients systemop quality is high accurate. High credit is high and high frequency.

19:03.98
Narayan
Right? right? right? again you know super interesting from all the insights that you are sharing in that Sense. So ah from a perspective of an end user that is looking at all of these things. Can you give a perspective on what were their day doing before for example with respect to their business processes and how much value you are creating for them. You know in their own markets and the kind of efficiencies that they are deriving based on the services that you guys are providing. So.

19:27.46
Gunjan
Yeah.

19:37.79
Gunjan
2 so ah, most of the companies right? They rely on reports data generated by the traditional market research companies. Ah, it could be. Ah, your original players like myself with are very big companies actually in China who provide a lot of market data. It could be companies like bloomberg ators etc who are s and p its who provides reports and companies rely on those information. So ah, most of the end users are already exposed to a lot of data sets which are coming in and using which they and they have a set processes which haran now our job is to provide a first task as a new player in the market I was to say you know? ah. Our data is accurate and ah we give ah our data. Um, ah much um, at a much higher frequency. So ah, most of our datasets that come out either come out say fifteen twenty days after a month is over or after a quarter is over while we provide data sets. Ah, either weekly or daily. So. It's a ah leading indicator of what is going to happen second. Ah we provide data irrespective of location. Ah now other market research companies some somebody strong in a particular ageography. And um, so it's all fragmented like um so but we provide information irrespective of which location you are interested in so and there are very important. Um, ah, interesting use cases for example in Southeast Asia countries like Vietnam. Ah, which is growing very fast in terms of industrial activities now a lot of information do not come out from that region and that region actually is starting to move the market commoity market. So our data is being lied up and to provide insights. On that region because no one else is reporting that those numbers then third ah we look at our data as more objective and scientifically derive than other players who rely on ground sources or connects. You might say to to provide those information. And the basic um way in which you derive data from ground sources also lends itself to various errors. Ah so these are 3 Usps that we provide to our clients now there are the.

22:26.59
Gunjan
Now there has been some early adopters. Obviously there is ah a pushback. No. Ah, there is a momentum they have to make changes in the and internal processes to no adopt a new kind of information, a faster kind of information which will require change so we have been working with some of the early adopters. Ah, ah, who ah have been using our data for various use cases from understanding say how much coal to sell in in the asian region. What what would be the need for coal in different areas of the asian region so they can um. Ah, ah, monitor or change of their supply chains accordingly and also ah derive some pricing information about what could be the basically pricing points at which they could sell then there's a particular client who has been using our data sets in for the southeast asian region. Um, to ah basically ah write do a lot of analysis on top of that and ah they have been selling those analysis to the larger industry which is the physical traders miners etc at at the cost. So ah, they have been able to earn a lot of revenue now it's difficult for us at this stage to assign a particular dollar value because it's still a little early as we go along I think this year um we should be able to and know what what would be the ah no, ah do i. Ah, for our datasets. So but we do believe they'll be strong and we we are working very closely with some of the very large ah companies in the sector and when we work closely with them I think ah will be able to. Know I learn a lot from them. Also you know how they have been using it and some of some some of them do not tell us how they are using it the data so we have to create those relationships to understand at what exactly at what not touch points. Those are it asset solve which problems for them. More precisely. The cell.

24:41.22
Narayan
And from the sources that you mentioned earlier you know you mentioned mois and lsat and maybe even sentinel and others did you ever try using indian satellite imagery or was it just too painful to get anything from India. So.

24:46.29
Gunjan
And.

24:55.91
Gunjan
Ah, no, we did try ah now. So ah, we ah know actually ah we had we have a fairly good relationship with is Israel. So no, ah, we had multiple discussions with andnna and rc. On our algorithms also with the scientists at an rc so on the data part. Um, we wanted to use and see no if we could integrate data from Israel also in our workflow. So ah. I would say right now I think the process is not as straightforward or easy as as working with say some of the other commercial satellite providers. Ah so ah, for example, we just wanted to check for two months you know what's the frequency of data that we would get so. I would say we receive data for one week then we didn't just receive data for three weeks then again we receive data hundred Fourth three probably so those kind of issues are there. What I think is ah but they are probably making ah efforts to overcome those situations. But right now. Yeah so I think there's a lot of catch up to do. But regard to no if compared to other commercial providers or in isa I think isad does a establish job of not providing datasets and handholding startups or handholding companies in multiple different use cases. I think yeah's that's the way to go. But I think that's a step in the right direction of for you. So also.

26:38.43
Narayan
Yeah, absolutely and from perspective of you know customers and adoption of these kinds of economic indicators. You know is this something that. You see starting out in India at the enterprise level or are we still at a very very early stage where people are not really valuing this kind of thing or is it only segmented to agriculture or places like that or also in other industries like the industries that you guys are working in metals and others. Is just too early or is it that you have to educate the market a lot in this environment.

27:16.30
Gunjan
So ah, it's not just not India actually the enter industry we had to educate and ah for them. Ah day don't care whether the data is coming from satellites or from other sources as long as we give them accurate data at a faster interval. So that's so and that's all that is at the backend. Whatever is there I don't think that is important for them this they know the cap app capability and they know what use cases that data can so if for them is just a piece of data. Ah right? So ah now ah say in the metals and mining sector. Ah.

27:55.21
Gunjan
Particularly in the Ferris industry that we started of the market is now the commodities market generally no is driven by China because ah they are the largest producer consumer of most of the commodities. So naturally it know natural for us to start off in the no Singapore region where which is the trading hub so as we got more feedback more clients in that region now. Ah, like late last year we started talking to some of the indian companies in the metals and mining industry so there is a lot of interest intrigue I would say to see what they could do with this data. Ah there are meetings lined up this year this and couple of January in February. Very. But some of the largest companies in India in the metals and mining sector and let's see if we can partner with some of them so I do not be a seaki. It could be a problem of say India versus the rest sort of but if it solves a problem in them companies or adopted e earlier. Is just that the market is larger outside but indian companies would definitely if it solves a problem for them it it definitely adopt. It.

29:16.91
Narayan
Right? And one of the things that I often think about is how do we encourage other founders to follow the same journey that you have where they discover something or we can support. You know so in some initiative. Having ah, kind of a geospatial acceleration model or an incubation model in the country. There are many of these kinds of models available in Europe and us and many other places around the world. Do you have based on your experience so far any insights into what can be done.

29:35.93
Gunjan
Um, yeah, yeah, yeah, yeah, yeah, yeah, is.

29:52.23
Narayan
But respect to a geospatial incubator or accelerator specifically in India and.

29:53.21
Gunjan
Right? right? right? Yeah yeah, great point actually so we always thought isa take the example of isad have been a great job I think to working with companies so we always and we some of our contemporaries in the european e. Ah, region because it's it's they is so easy for them to get grants to work on ah to work on specific problem statements or on specific areas using geospatial data and ei is always supporting them Isa Assign Isa who works with them to know develop new methodologies. So if you say come. There are some companies like no kiros or who have been strong partners with Issa and is spread across know there's are two c companies in Uk a couple in France. Ah. Who are partners at Lisa and they work on multiple business cases not on us on the remote sensing side I think a similar approach can be is is very essential in india either isro does that or we have another think tank or research body which works. With no entrepreneurs startups to figure out different areas. How this data can be applied to ah and I think ah ah and the financial support that could be provided at the initial stage regard. This. Areas are not no on the market where say mainstream venture capitalists come in very early. They'll and they'll like to ah see in abstractction or because first of what for our case, right? They're not in our space accelerators or space space. Ah venture capitals in India. Ah. Of them then also look at the sector that we have taken ah metals and mining. So. It's a combination 2 which is very difficult to get in India people who will have a thesis around investing in this to in a come in ah in ah in a industry combining space data with metals and mining sector. so so and and initial stages I think a lot of um, grant work or know providing small grants like 1520 lags thirty forty likes which will encourage people to work with on multiple use cases and also I think there is only possible. Ah, now say financial support second if the data is available easily right now. Would you expect people to again indian ah is so to give grants to organizations of startups which work with yeah which work with say data on Esa or Nasa.

32:40.15
Gunjan
That that isn't very prudent so you'll possibly believe you know people who explore is is row's datasets to come out with solutions. Um, so and there I think a lot of industry support is required around providing access to data and providing some. Grants at Thy initial procedures.

33:02.53
Narayan
Right? And from your perspective is there any like room for a lot of the mature enterprises in India to look at you know in-house geospatial innovation in some way or the other because one of the things that I've seen with. Many of the larger enterprises around the world today is they have these geospatial labs that they have now started to build in-house. It's been a trend for the last like 15 years Maybe you know at the end.

33:29.27
Gunjan
Death.

33:34.75
Narayan
I've not not really seen many of this occur in enterprises within India for example, large-s scale enterprises. So.

33:38.10
Gunjan
Yeah, yeah, yeah, definitely this this I think in the earlier answers I afraid I have added this I think a collaboration between is 0 large enterprises and ah startup is required so ah and ah another problem statement that. Starting on the journey is finding the right andka clients that are and the first set of clients who will tell them can exactly. Ah, yeah, yeah, this really solves a problem for me so in India of course there are very large companies with assets wherever there is where there are companies with large physical assets I think geospatial data. And play a very big role and this companies can definitely definitely add a lot of value in terms of no sharing use cases and being the initial clients. Ah for startup. So. People working on this ah geospatial data sets.

34:34.60
Narayan
And how mature you think the whole geospatial industry is today with respect to you know the kind of work that you are doing where people don't really care if it is from satellite imagery as you said, but they really kill about the value right? But you know at the end.

34:45.12
Gunjan
No yeah, yeah.

34:51.17
Narayan
There's a lot of fascination about the technology itself and obviously you know people don't care how the technology is built. They look at the value that that particular technology is developing so are we at the very beginning of this geospatial revolution where people will discover this more and more or you think we're in the middle of this revolution today and you know.

34:51.62
Gunjan
Um, yeah, yeah, yeah.

35:11.10
Narayan
What is the scope of this industry as in. Do you think that every you know fortune 500 company would be using geospatial data in some way or the other are so.

35:11.73
Gunjan
Yeah, yeah.

35:19.44
Gunjan
Yeah, definitely I believe we are just at the beginning because a lot of I think physical ah infra is still getting built a lot of satellites are still going there. So ah now let's take an example of say 2 3 years back ah say at that point of time when a lot of companies started. Probably it was not even possible to give a lot of insights that we can do now. Ah data was not available. Ah, there was ah continuity of data is a big problem. So ah, if ah, those unless and otherwise those issues are taken care of. Around quality quantity. No continuityity of data. Ah, it's very difficult for businesses to adopt those insights deliver a few businesses which will adopt those inside now as we mature towards an ecosystem where the problem of data is being taken care of. As we see more launches more satellites going there and the type of data being taken care of ah ah so it's not just we need. We don't need only optical images or more optical data. We need a mix of optical thermal radar saa data sets hyperspetual different kinds of data sets which go together into into an engine to solve a problem. It's an insight it I think from my experience. Also we have seen. It's just not possible to work with 1 type of data set to solve up customer pain point we have to mix in different kinds of dataset from multiple providers to solve any particular pain points to maintain that quality and continuity of information that we give up. So ah, when that happens I strongly believe every big company could have would be able to derive very useful information based on space data. So I mean ah when it really becomes feasible to look at every point of the earth every day at the? Um, no. At a cost which is which makes sense for a business to invest and that that probably would be a tipping point when I think every big company or every company should be able to know bring a. Make their businesses more efficient are using geospatial information and our platform actually is if we think of our platform that we are building as towards making that possible because I know let's say how we are building. Our platform is any user. Ah, can select any asset. Ah, for right now. It's restricted to the metal man sector. So no you you just upload the location of any asset and you should be able to get every important information regarding that asset be it.

38:04.13
Gunjan
If you upload a location of a steel will you should be able to get production Inventory emissions data for a steelment very instantaneously so that should be the ah debt we believe should be the future as you over.

38:15.67
Narayan
So right? and 1 of the other again, very interesting pointers that I would love to have you talk about? is you know when I look at the human resource base in India we have an extremely large pool of very talented engineers who are building all sorts of things when it.

38:18.72
Gunjan
Um, yeah.

38:30.34
Gunjan
Um, yeah, no yeah.

38:34.98
Narayan
Respect to information technology right? But there's very few of them who have discovered that they can apply some of those principles and jump into the whole satellite industry like you have for example, right? So from that perspective. Do you think that.

38:40.53
Gunjan
Um, yeah, yeah, yeah, yeah.

38:52.87
Narayan
You know this industry is still at the edge and people are not really discovering this the whole you know one of the things that I always wonder is why aren't more and more people who are in infos or Tcs or you know vippro or wherever.

38:56.10
Gunjan
Um, yeah.

39:06.28
Narayan
Ah, you know who have 1520 years of experience and understand everything from cloud architectures to enterprise software and everything else and know how that particular industry works in other sectors are not jumping into this sector saying that I know something that I can build on top of this and you know provide some enterprise value to all of these guys.

39:06.37
Gunjan
Um, lately. Um, right.

39:24.89
Gunjan
Lately? Yeah, yeah, yeah, why.

39:24.92
Narayan
I haven't really seen that happen yet in India for a large extent. So why do you think this exists and you know do you see of course an opportunity in all of that.

39:31.26
Gunjan
Of course, there's an opportunity I think the primary reason is ah awareness I didn't know about is and until I bumped into a problem statement which required me to look at know. Ah ah, satellite images. So I think it just ha. Awareness ah for now even now. Ah when I talk to people when I talk to my friends or when I say satellite images they think it's oh it's it's it must be very very difficult to do. How do you get satellite images. Ah how? Well the Fascinat fascination and a fascinator satellite image. Ah, for us. It's just an image that comes in so I think that's the barrier to entry because people think is still ah set in India people think still think satellite space is something only the governments can do and ah now if people are not aware of the data. They they won't be able to imagine the business cases that can be solved using. Ah this new type of information that is very widely available right now. Ah now another reason could be in India ah, ah, some people in ii sector right. Ah, what are the kind of problems that they are exposed to for example, metals and mining industry I don't I don't will I don't think 80% of the people in diet sector would be exposed to the metals and mining industry or to the crude industry or to the agree in the agree industry. Ah, definitely because I think India is a predomintly agree economy and ah in their background parents when and somebody would be in the e sector so they understand Aris sector so but people do not understand the crude sector the and metalx and mining sector or the shipping economy. No I did not I was not exposed to this economy and ah. When I and think now say the marine industry which is a trillion dollar industry move goods. Ah with our economisties cannot exist without them and in in our education or our upbringing. We were never never exposed to those industries and I think those industries are still now predominantly. Um. European ah ah I would say ah so yeah, so is all this Ah no ah reasons combined I would say you know we have still not been able to ah have more companies work on more use cases regarding ah using geospatial data. So.

42:05.51
Narayan
And what are the challenges like for you guys now is it. You know hiring is it. You know something else at the moment I always wonder if hiring is a big problem in India because you know at the end of the day. One of the things that I see is a large pool of people who want to get into the industry. But.

42:16.49
Gunjan
Yeah.

42:24.12
Narayan
May may not have the relevant skills. For example.

42:24.13
Gunjan
Um, yeah, yeah, yeah, yeah, so ah, hiring for us I mean fortunately it has it had not been a big problem because ah ah we have ah tapped into sounder specialized institutes like ity bomba. For not hiring specifically for remote sensing rules and it has worked out very well for us. Ah so ah, now it's not difficult for to for any sharp say software developer or a data science. Scientists to move into a remote sensing rule because the basic principles remain the same say ah a camera image versus a satellite images. There are certain differences but but at the same time with all physics that they are exposed to so I don't think ah that should be a problem. Ah. So um I think for us it has the biggest problem has been reaching out to new clients because we are based in India and most of our clients are brought so it's easier for like with exist existing clients. It's easier to. Maintain the relationships and no sell more services. But I think what's been tougher is to convert new clients when ah we are all remote and in an industry like this where everything is new for the and client customer. Also so I think. That where that that that has been a big disadvantage because none of us have been um ill to travel to meet clients either. They are in the lockdown or we are in lockdown. So no, it's just not ah worked out where we could so say go to Singapore stay there for two months now ah so meet everyone understand the problem statement at once and come back and deliver that so that's not been able to. We have not been able to do that here of hopefully in the near future this year we should be a traveling lot more um to meet our new customers so that is problem statement number 1 I think a second has been. I not say a big problem but the the I think the ah ah ah, the ecosystem around venture funding is also getting built in India so we have been able to raise some funds earlier. So as as you as we all know. So. There are some specialized funds who will only look at this sector so it is still not mainstream yet. Not many come not many venture capitals will have a thesis around our industry. So yeah, so we have a limited pool to work with when we come to space applications.

45:17.60
Gunjan
Course now Space applications will combine what space data plus the industry where we are solving those problems sequence. So it grabs even more her remote or the pie gets even more smaller in that case.

45:29.23
Narayan
Yeah I mean again, super interesting. All the insights that you are sharing from you know from your experiences so far and I guess you know you guys based in Mumbai has some advantage to where you are because a lot of the.

45:35.50
Gunjan
Um, yeah, yeah.

45:46.26
Gunjan
Um, yeah, yeah, yeah, we are here. Yeah, great trip.

45:47.19
Narayan
Economic indicators and the financial industry around economic and you know indicators are all around. You guys may be exploited in 1 way or the other have you got a sense of how deeply are financial markets in India using any of these kinds of satellite-based indicators or space-based indicators or. Is it that I know that for example, you know, yeah ubs and many other bankers and and so on so they use a lot of these kind of indicators to make a lot of decisions I'm not really seeing that a lot in India yet I know that power space and maybe a couple of others are working on this front. For example, is it just too early or the are the trades too small.

46:12.27
Gunjan
Who right right.

46:26.70
Narayan
Force people to pay for this kind of information.

46:27.38
Gunjan
Um, it's ah it's both I would say um now ah um India is primarily a equity market I believe people invest in equities while ah. When we see when our well are and datasets can be primarily used for either macro investing no because metals mining you combine everything together. You'll get a good picture of the macro economy or into commodative trading now commodity trading if the financial industry for commodity trading is not evolved in India. For financial for financial ah in the financial market in for collaboratative sector. It's either. Ah London um, no, where they have a where whether is sister London or not America or Singapore. So those are the primary markets for com trading. So the market is itself around finance in the financial industry for trading on commodities is limited in India now ah around I think around sector also as you mentioned the company is doing it I don't think in India the financial sector. For agri commodity trading is very deep ah ah compared to North America so it's just not the space data. It's the markets themselves which are limited in India which now yeah, which limits the usage of sophisticated technologies now. Because if you use this data. You also have to generate and that alpha or that ah roi from this information. Ah which I don't ah which at this stage is probably not sustainable in the markets that indian financial industry works in. But if however, if somebody say gives information around say the automobile market auto sector. I I think those kind of informations can be adopted in India. Yeah, how much cars are being are being manufactured each month at different manufacturing units I think those other information that can be used by the indian financial industry. Um, yeah, so yeah so those other reasons I think the adoption in India has been lower in the financial industry search.

48:42.45
Narayan
And 1 of the emerging topics around the world is this esg the environmental you know, social and governance perspectives for a lot of the sustainability of resources and climate change and others and.

48:46.41
Gunjan
Um, yeah, yeah, yeah, um.

48:54.49
Gunjan
Right? right.

48:57.88
Narayan
From your experiences so far in developing products for this metals and mining industry. How deep has this concept of Vsg gotten into the enterprises in the metals and mining space and are there anything beyond the economic indicators that they are trying to evolve to.

49:00.84
Gunjan
Oh.

49:08.67
Gunjan
Are. Um, yeah.

49:16.69
Narayan
You know, fit that into this climate change or yeah, esd framework.

49:18.34
Gunjan
Yeah, definitely I think the metals and mining industry. Ah so I'll give examples some of the resource companies mining companies. So they very thoroughly track their entire supply chain. It does not scope on emissions they and they monitor scope to scope three. Also. How much so if somebody is mining ironode. They'll also ah monitor which ships are not ah carrying those goods or commodities and what's the efficiency level of the say steel mills which are using those ah those. Ah, raw materials. So the entiremet mining industry is I would say is very aware ah of the eg space and efforts are being made and ah to improve the hs course and I think here satellite data would play a very key role for them. Ah. So there are conversations happening around ah can satellite data help them trackk asset level emissions. So right now we understand proper me o seo or go set can ah we can provide emissions but at a very large area scale. But. And the problem the ah the underlying problem statement is can ah no small areas. So small assets can also be tracked now. So ah, we are working on certain interesting products on that. In fact, we have observed using thermal imagery. Ah, we have been able to. Ah, no correlate or closely match the ah um, the declared carbon dixide equivalent equivalent numbers from many steel mills in ah in certain geographies. so yeah so I think Hh and the h e ah the environmental sector would be a big ah market. Um, for for remote sensing companies for satellite companies and remote sensing companies as I think ah and I think a lot has a lot will evolve as policy makers ah financial institutes. Everyone. Get around to exactly what snapffs to be tracked exactly what needs to be done right now. There are a lot of discussions going on everywhere. We hear a lot of chat. But I think something concrete has to come out. You know what? exactly other metrics is that need to be ah tracked. The asiotrics themselves are not I think standardized yet or there's no strict regulations around how they are reported everyone msci and does it. But yeah, there is no standardized way of doing it I think set and on the environmental side at least satellite data can I think play a very big big fact very very big role

52:10.52
Gunjan
Ah, know in understanding both environmental impacts of companies or assets as well as understanding how much at risk are those Ah Industries units assets are 2 climate changes. So we are looking at both this angles very closely right Now. Actually.

52:33.87
Narayan
and and I would love for you to confirm something that I have as ah as a theory in my mind which is that I feel like India is going to be a great place for this kind of geospatial companies to come out in the next you know few years especially because of the talent pool and.

52:39.28
Gunjan
Um, ah.

52:48.75
Gunjan
Um.

52:53.67
Narayan
You know all the itd infrastructure that we have and a lot of the access that we have alongside the you know cost of infrastructure decreasing and and the talent pool of course is much less expensive than compared to us and other places. So.

52:53.84
Gunjan
Try to.

53:00.13
Gunjan
They look. Yeah.

53:06.54
Gunjan
Um, yeah, yeah.

53:09.59
Narayan
Also feel like the cultural understanding of problems may be much easier for people in India to do and solve it at certain price points that many other people may not be able to to replicate in places like us or ah Europe for example, right? So I feel like you know, a lot of the solutions that will probably get built in India.

53:21.50
Gunjan
Great.

53:28.81
Narayan
We'll be able to easily scale outside in places like Africa or Southeast Asia or otherwise as well and and so on so what are your thoughts around all of this. So.

53:31.69
Gunjan
Right? life right? I think I can and agree with your thoughts actually on this so ah, building from India gives us a huge ah no advantage in terms of cost. Definitely. Um, ah and ah now if you can build from India at a price point which is not just competitive with but much lower to other other european nations. So but who are we competing here. It is is is either european or not american companies and. I think the economists there are very different. The price points are much much higher for the same product. We can sell at say fifth of the cost just because our economies work well from here. So it's much much easier for us to scale then whatever we built for India. As rightly pointed out weak can scale to southeast asian south asian nations to know african nations. Um, who probably have the same kind of problem statements to what? no we have when we first started out with a problem statement of understanding. Ah, no say per capita income in India as who so in us, you don't need that service because you already have ix exercises and all where they trek pincoat level asset level. So but it's its it is comparable to our Asian Economy Southeast asian Economy South asian economy african economy where we still rely on manual service done every every five years every 10 years where no growths androsome um money is spent to do this service which I think can be done much more efficiently much more at. Much higher frequency using satellite images or using other alternative data sources and it will fit with all the the economies that you just mentioned I think yeah, it can work very well building from India.

55:35.77
Narayan
So right? So thank you so much again for taking the time we already spent almost an hour and it's been really fascinating to hear all the things that you're sharing. Maybe 1 last final question before I'll let you go.

55:39.40
Gunjan
Um, yeah.

55:45.11
Gunjan
Um, Shasha shaha.

55:49.47
Narayan
Where do you see yourself and the team at Thatya in the next. Let's say 5 to 10 years time you.

55:53.72
Gunjan
Ah, so I think in the next five. So let's let me speak about 5 years only got 10 years is too long ah time horizon than I think to predict. Ah so in 5 years we should be able to map all key assets across the world. Ah, in the metals and mining industry and will probably be ah in another ancillary industry other than after having solved for the metals and mining industry and ah we should ah have ah. Products around climate risk and on and so on sustainability. So these are 3 products that I am volish on for the next five years ah and are ah regarding the team I think India would always be that development center. Ah, but I think it is essential ah for us to have a european ah presence and a not american presence also along along yes, along with Singapore and e along with the presence in Singapore um, so yes, so so multiple countries. Metals and mining industry solve for checking of supply chains and a product on sustainability and climate risk. So yeah, that's what we tend to intend to solve in the next five years

57:22.87
Narayan
So right? Thank you very much Kunjun I think this has been very educative for me and then I'm sure that the audience will also love hearing this particular conversation as well. So thank you again for taking so much time and good luck to the entire team. Ah, Tatya and I'm always going to be a keen observer and learning from your developments and products.

57:43.19
Gunjan
Thank you so much Noran and thank you for giving the opportunity to share my thoughts and yeah i'llll'll'll I'll I'll keep in touch and yeah and we share will discuss I think how things are are shaping up at our end. Ah, frequently.