ChainTech Demo Day, May 2023

All right, while we're still getting people on board, I want to again welcome all of you to that first Gentech Demo Day series, you know, together with— we organized together with the Global Superchain Council and Gentech.net. Again, as I explained earlier, you know, just to give you some guidelines, this is going to be an interactive session. Each founder is going to be introducing their solution, giving you a live demo of what they do, and they will also have time for your questions. So, you know, you can ask questions at any time using the Q&A tab located at the bottom of the screen. And then, you know, after each demo, we will have about 10 to 12 minutes for Q&A, and then you can then ask a question and get answers directly from a founder. I also want to, you know, inform you that this session is being recorded, so if you want to rewatch it or maybe share the recording with your colleagues or friends. We'll be sharing the link after the live event. And also, you know, if you want to, we'd love to introduce yourself in the chat. You know, the chat is open. You can tell us who you are, where you come from, and what you do. And then we'll be, you know, this way we can then just stay connected with you. I want to again remind everyone not to ask questions using the chat. Please make sure to use the Q&A tab located at the bottom of your Zoom screen. But if there is a question in the chat, we will try to push it back to the founder. But again, please make use of the Q&A tab. All right, so without further ado, I'd like to bring on our first founder joining us, you know, Venu Goutlapalli from Tag&Track, who's joining us very late from the US. Hi, Venu, how are you?

Hi, Max. Thanks for having us on this Demo Day. Uh, it's a pleasure, and, uh, we're doing very well, and we've been very busy, uh, last few months trying to get this, uh, uh, deployed in the field. So that's what we're, uh, nowadays we're, uh, going after— deployment. Okay.

All right, go ahead. Welcome.

Thank you.

Um, let me share my, uh, screen and we can then go Do you guys see our presentation?

Yeah, you can go full screen.

Yeah, let me— all right.

All good.

Hello everyone, I'm Vinay Guttapalli. I'm the CEO and co-founder for Tag&Track. We are headquartered in San Diego, and it's my pleasure to be on the Demo Day platform today. And I have 3 slides that we walk through who we are and what we're doing. Then I'll jump into our platform and we'll also show how our smart labels work. So as I said, we're a full-stack IoT traceability and visibility platform as a service company. We're a hardware-enabled SaaS platform, so we build a full-stack IoT from sensors all the way to the edge infrastructure, to the cloud compute for data acquisition. We have a global presence. We are in India, we are in Singapore, Germany, San Diego, and Silicon Valley. And we have raised so far around $11.5 million in funding, and we just closed Series A, $10 million, and currently we are at $1 million revenue. generated. And then our lead investors, or Dell Capital is a lead investor, and we took 2 strategics. One is Global Health Innovation Fund from Merck Pharmaceuticals and Aerosafe Global, which is a packaging company. And we primarily focus on pharmaceutical cold chain, third-party logistics and freight forwarders, and healthcare and enterprise asset tracking. So that's a summary of who we are and what we're doing. And in terms of what we're trying to really solve is if you took a look at any supply chain end-to-end, you have this fragmented data collection at the manufacturing side. Somebody is using RFID, somebody's using barcodes, then they get packaged, and then they call the 3PL or a freight forwarder to move the goods. Then the, the 3PL will come, and then they'll either use an offline data type way to provide visibility, or sometimes they use some kind of a tracking device to provide visibility. Once he delivers to the destination, his job is done, and then again, you'd have no visibility continuing forward. So what we're doing is we're trying to solve that fragmented data acquisition for supply chain. and make it seamless from manufacturing to the retail, all the way to the last mile. We try to provide visibility end-to-end. So part of our stack is a, is a device layer, which is the bottom layer. You can see a lot of devices there with some infrastructure gateways and handhelds that we use for reading and mapping barcodes to the trackers. Then we have an infrastructure layer where all the data gets pushed through. Either it's a carrier connection activity or through wireless and cellular gateway infrastructure. Then that data gets pushed to the— our IoT traceability platform. It's called Relativity, and real-time visibility and traceability platform. And that's where all the data gets aggregated and we process the, the data, whether it is inventory cycles, warehouse inbound, outbound, whether it is transportation visibility, everything gets managed there. And from there we push that to our next layer, which is a continuity layer, which is an end-to-end data orchestration platform for supply chain where all your partners of your supply chain can be onboarded onto the same multi-tenant platform and everybody can see their, their part of the supply chain data put there for visibility. So that's our complete stack, how we implemented it. And this kind of solves the data fragmentation, and it makes it turnkey for the supply chain vendors, either it's an OEM or for the 3PLs, to offer this as a service. And we take care of the, the entire deployment for them, and it's very easy to deploy IoT with us. And one of our differentiation is that most of our sensors, they come as labels. So we take any device that is previously used as a tracking device or a big bulky device, we converted that into a smart label. And these smart labels can go through the printers and you can print the same information that you're printing before, but then you don't need to scan it again because when it's printing, the camera attached to the printer maps the data what you printed for your shipping items, then the sensor embedded into the, the label itself is mapped right when it's getting printed. Once it's stuck to the pallet, now the label is transmitting actively its ID, its location, its context of where it is and where it's going. So you get a lot of intelligence without having to scan again these barcode labels. So we're kind of next-generation RFID. You can think of it like that, more intelligent more sensors. It can temperature detect, it can do tear detection, it can do tamper detection, motion detection. So we can do multiple use cases, uh, um, with, with these labels because we can embed different sensors into these labels. So what I'm gonna do is I'm gonna play, um, our video of how our labels work. So I'll give some background commentary as we, uh, play like how we're solving the problem. This is our portal, and once you land into our portal, all the warehouse printers— these are Zebra 411 type warehouse printers where you, you're printing your shipping labels, like 4x6 shipping labels— we can connect those printers to our cloud platform, and you can pretty much create different IDs of Labels. And once these labels are created, you can just say print. And that's the label we just created with the shipping ID and where it's going. And then once he says print, that label is getting fed through a Zebra printer. It's the same warehouse label printer, but we have enhanced it with a camera and a full cloud connectivity. So we converted this regular printer into an IoT printer with vision attachment. So as it's printing, it's printed the, uh, the, the shipping information, and then you can see in the bottom there's a QR code. Now all these codes are mapped. Once that mapping is done, you don't need to scan again. So now the, uh, the customer is sticking it to a pallet that is going, uh, getting shipped, and once that, that is shipped onto a pallet, stick onto the pallet, it's now actively transmitting what it is doing. Now you get this end-to-end data without having to scan this again. When it gets to a truck, we can track the truck. When it gets to the warehouse, automatically the warehouse gateways pick it up, and then we can do the full inventory of what came in and what was the inbound and what is the outbound. So that's what we, we're doing in terms of the label and the data connectivity and how we manage the entire inventory cycling. So with that, what I'll do is now let me go through our portal, how we manage that data. So this is our portal. So in our portal, you can see that there are multiple warehouses worldwide, and each warehouse kind of picks up the whatever inventory that we're supporting. Now in the dashboard, we kind of show the— what are— how many total assets have been attached with smart labels, and out of the ones that are attached, how many are reported through warehouses and how many are in transit. So you get the full real-time visibility of your overall inbound and outbound. And then every warehouse we can create into zones. And then here we're showing you the inventory cycles that are happening at a zone level. So we can create— you take a warehouse and we can convert that into inbound zone and outbound zone, and then you can pretty much get real-time inventory of that. The other thing we can do is Also, there is sites, and within sites we can create, you know, the inventory reporting that we can show. So some of that I can show right now. Like, you know, this is one of our warehouse, and we can— we have like 3 readers. And currently, if you see this, pull the real-time data from our warehouse. Around 1,200 assets are in this warehouse and it just pulled it real time in less than 2 seconds. We got the full inventory cycle from the warehouse. Whereas if you have barcodes only, paper barcodes, then you have to send somebody and then they're doing the barcode visual scan of every pallet, every asset in the warehouse. Instead, you can automate the whole process. So that's one of our key implementation that we did. And the other thing that we do is also— let me see how we can— so we also do the transportation side of the visibility. So here what we can do is when you are moving multiple shipments, whether it's an LTL or a full truckload or a light truckload, We can create with the smart labels, we can create purchase orders with a bill of lading printed on it. Then you have each order combined with how many packages, and we can then track this in real time. So here we're showing like how we did a delivery of a shipment. So here is every shipment that we are— the session expired. Yeah, so here you can see that there are multiple shipments that are currently being either shipped, they're delivered, or they're in the process of delivery. So I'll take this example shipment that's already delivered. So if I click on that, it kind of shows that this was a shipment with a bill of lading number of this, and it was a purchase order number 1. It had 2 packages. And it was going from this warehouse to the destination warehouse is Oklahoma, and it was shipped from the current location, and the source address is Tennessee. So that's where it's going from source Tennessee is where the fulfillment center is, and the destination Oklahoma it's like a Walmart center. I think that's where we're delivering these, and we then capture every. every use case. As the shipment is progressing, we can provide the what happened to the shipment. We can also look at how many packages are there in the shipment. So it can tell you like there are 2 shipments in this that went with this shipment. And then it also gives you like a FedEx, like more real-time data, how much time it took in transit, where it was waiting. when the shipment is prepared, when it is ready to go into the truck, when it got delivered. So these are some of the things that we can do with these labels, and all of this is automated end-to-end. So that's the summary of like what we're doing. And at this point, I would, you know, leave the floor open, and I can also show some of our devices. If you guys have any questions, I can answer those. Thank you.

Yeah, Vinu, if you would like to maybe stop your screen sharing and show us the device very quickly.

Yeah, let me do the stop share. All right. So some of our labels, they look something like these. So they come in a Z-fold format, and then they're a 4 by 6 label. So they come blank. When we supply to you in a Z-fold stack, and they come with this barcode printed like a QR code. That's the ID of the sensor inside. And if you, if you print it, then you print the actual shipping information. Then what I showed was that, that, that information and these 2 shipping barcodes are mapped. Once these are mapped, the printer then will cut this label to activate them. So there's a mechanical cutter that cuts it, and then that's the activation circuit. As soon as it cuts it, or you can even tear them. So if you tear here, then automatically this device gets activated. So if I just do that, now the, the label is activated. This is actively transmitting Bluetooth, and now because your smartphone is also Bluetooth enabled, You don't need an RFID gun. Any smartphone in the warehouse or a tablet now can read this label from 100 meters distance. And all the data now is coming automatically. And one thing we also do is we provide these gateways at the warehouses. So this is a multimodal gateway, and this has cellular, Ethernet, Wi-Fi, based on what you allow, the customer allows us to connect to the connectivity. They can then be installed at the inbounding, outbounding, and also in the aisles. So we can give the full warehouse coverage. With 5, 10 gateways, I can cover like 10,000, 20,000 square foot warehouse. And once these are installed and these tags, when they're on the pallets, automatically they're chirping every 30 seconds. So every 30 seconds, this is advertising who I am, and what my metadata I'm carrying with me, the pallet information. And then that information is almost going to the backend cloud at real time. So you don't need inventory cycles, you don't need to send people to figure out where your pallets are. I can provide almost like a zone-based location, or I can even do sub-meter accuracy location where your pallet is in a given place. So we not only provide these type of label formats, we also have different formats. This is our temperature logging device. This is pharma-grade temperature logging device that you can attach to the packages and parcels. We also have much more tinier for smaller use cases. This is our smallest BLE sticker, and this can run for a year, and it can also measure temperature. This is mainly used for much smaller packages at a SKU level, if you still want to put little items to track smaller items, then you can use that. And then for pallet level, you can use the labels. And we also have this type of tags, and these are one of the lowest industry low-cost tags. A typical Bluetooth tag usually costs around $10 to $14 based on what sensors you put. But we can make these at $4 or less based on the volume. Sometimes we can even get them to $2 or less. Our goal is in millions, the labels, we want to produce them at $1 per pallet. So that's our— how we are designing and manufacturing them. So we design everything in-house. We manufacture all this in-house.

So, okay, thank you. Thanks a lot, Vinay. All right, I want to remind everyone that you guys can ask questions. I see some questions coming in already, but go ahead if you have any questions for Vinay about tag and track solutions. You can just hit the Q&A button and then we'll take those questions right away. Just very quickly, Vinay, I mean, before we take those questions from the audience, a few years back we talked, we had that you know, that big buzz around RFID. Is your technology better than RFID, and what's the difference?

So RFID is, again, so the main thing is like how you collect the data. RFID is a pull data. Both barcode and RFID is passive because they're not transmitting all the time. They're like a dumb tag unless you point something at it to pull the data out. They don't respond. So it's always a pull data. So you still need to send somebody manually to go pull the data, whereas the Bluetooth and the cellular labels we make, this is push data. That means the data is always getting pushed to the cloud so that you don't need people go and try to figure out, you know, how many pallets I have, what is my pallet doing, what is my temperature of my pallet if it is perishable goods. We can automate that whole process, right? So for example, An average air logistic package or a parcel needs 17 scans if it's doing international shipment. If we put the smart label, you don't— you only need one scan in the beginning. That's also we automate in the printer. So if you do that one scan, that's all we need. After that, everywhere it's captured either through smartphone or through gateways, we can read that data. and push that into the cloud. So it's the— it gives you more automation capabilities.

Okay. We have a question from Jan who's asking, how long do the ID transmit?

So we have— it's a printed battery. So each one has different capabilities. For example, this can last at 30-second interval up to 4 years, 3 to 4 years. This at a 10-second interval can last for a year. So this little sticker, this is a temperature logger. This runs for 1.5 years at 30-second, 15-second recording. We record every 15 seconds. The labels, we come with the 2 different capabilities. One is 3 months, other is 6 months. So it depends on what type of lifecycle you want on a given label tracking where you are applying and how long you want to track it. If it is air logistics, most of the time you only need it for a week because most of the parcels get delivered within 7 days. But if it's an ocean shipment, then we would like to keep the label active for 18 weeks or up to 6 months if we want to consider end-to-end.

Okay, another question. Jan is asking, so do you need gateways in the trucks in order to have Events in transit?

Yeah, we can do it 2 ways. One, if the truck driver can run our app, we can collect through that. Or we also have gateways. This is global gateway. It can work anywhere in the world. And you can put this on the truck or on the master pallet. So let's say you're doing 20-pallet shipment, then you put all the label, everything, and then you just put this, the bill of lading, put this into the COVID of the bill of lading, and then you, you track your entire shipment. So we have multiple options to solve that infrastructure problem. We have a truck trailer gateway, we have an application that we can run in the warehouses. We can also pull the data through Cisco Meraki, Juniper Mist, Aruba Network, Extreme networks. If you have your warehouse has multiple Wi-Fi networks, we should be able to pull the data through them. We don't even need to install our gateways there.

Can you track sea shipment?

Yes, we can do ocean shipment tracking. We can do road, we can do air. So all 3.

Okay.

All right.

So another question asking— this is Ming asking, where is the data stored?

Um, so the data is stored in the, in the cloud. Um, it— everything goes to the cloud through the infrastructure. And if the gateway doesn't have connectivity, sometimes we record on the device, and the first opportunity we get, we upload into the cloud through the cloud, through the backhaul, whether it is 5G networks or it's Wi-Fi backhaul. Or it is the smartphone that the driver has, or the warehouse manager has a Zebra-type reader. I can do an inventory count and I can upload that into the cloud. So there are multiple ways to solve that problem, and we provide end-to-end.

Another question from Miong is asking, is a battery inside the label?

Yes, there is a printed battery. We print the battery along with the electronics, so the whole thing is assembled as a With the printer battery. Yes, there is a printer battery inside.

Okay. Another question also from Akhilesh is asking, is Bluetooth Beacon capable to track a vehicle on its own with no other hardware required?

Bluetooth, it can— of course, Bluetooth always needs an infrastructure to push the data to the cloud. So it will need either a truck driver's app or we give you a truck gateway that you can install. And it's a plug-and-play gateway that runs for 3, 4 years. So we have, we provide that infrastructure to read.

Okay. Question from John. What is your e-waste rating for sustainability and recycling?

E-waste rating. So one, one thing is that this battery, we're not using any rare earth metals or lithium. This battery is zinc magnesium based. And you can dispose these off with no problem. So the other material is same as your paper labels. Whatever you're using labeling, the rest of the material is same. Whatever is currently used, that is what we're using too.

Okay, question from Loknath. How does this integrate with applications like SAP Warehouse Management, Order Management, update the status on the delivery document or order documents?

So we have already finished the integration with SAP on the WMS side. And then we did TMS integration with DHL. We did the TMS integration with Kunnianagel. And I mean, it's all API-first approach. We don't write the applications mainly. Whatever I showed as an application, these are all demo applications. Sometimes customers use them. Sometimes we push all the data to the freight forwarder or 3PL's TMS systems or their WMS systems. So it's all API-first implementation. Everything can be integrated through an API.

You mentioned a few times the 3PLs that you work with. I mean, the question would be, are 3PLs supposed to do this? You know, I'm a bit surprised. So are you saying that they're doing this manually with barcode scanning, and this technology is actually disrupting the way they handle their shipment right now?

That's correct. So we published a white paper with the freight forwarder Kunenagel, and what we did is that we did an air logistics digital transformation project for them. So they move around 10 million air parcels every year, and if you take that 10 million parcels and if you multiply that with 17 average barcode scans per parcel. You can imagine how complex their operation is, and they're looking for ways to automate. And we did a 10,000-label pilot with them, and we have published a white paper. And if anybody interested, you can go to our LinkedIn, it's published there. It's also on our website, and I can, I can send you guys a link if you want to read through that where we publish the results and like how we did, how we deployed the solution for, for this type of digitization.

Okay, we have another question, quick, quick one before we wrap up. What is the Bluetooth technology used? The reason for asking is to understand the distance.

Yeah, the Bluetooth currently we're using is 5.1. It has line of sight is 100 meters. And if you want long, a little bit more long distance, we can go at max of 2.5 dB power. We can go to almost 150 meters line of sight, and non-line of sight is a 60-meter range. So we can pretty much cover half the warehouse distance with this.

Okay.

All right. I think we have a comment from Loknath as well, who's asking— who's saying use cases on the sales order delivery. There are multiple dates, but status is not updated. You can say it, I know what you're coming over here.

Yeah, so the main use cases that we're solving with this is one, as I said, the Kühne+ Nagel's air logistic transformation. The other use cases we did is one, we did lost in transit because the electronic high-value goods are getting either lost or have a theft problem. So we catch and report within 15 minutes of the shipment missing. We can alert the vendor that, hey, the shipment is missing. So we're also detecting the split shipments. Let's say you do an LTL and then you're supposed to send 20 packages together. And instead of 20, only 10 arrive because in the middle of the transit you split half somewhere and half are going somewhere. We're catching that. We do a lot of warehouse automation with these. We are doing multi-warehouse automation for all inbounding, inventory cycling, outbound, and the whole warehouse automation can be done with this.

Okay, well, with that, thank you very much, Venu, for sharing again what you guys are doing at Tag&Track. It was very interesting, lots of questions. And again, if you want, after this, during the next demo, leave you contact details, or maybe Steven, you can also leave a link to that white paper that you guys published so people can further understand your solution. Thanks a lot, Vinod.

Yeah, thanks guys. It was a pleasure having me here and I enjoyed presenting our technology today.

All right, moving on. We have our second founder who's on standby, who's been listening very carefully and is ready to go. I would like to introduce you to Gaurav Bhubna, who's from Next Billion AI. Hi, welcome, Gaurav, and this is your time. Go ahead.

Yeah, thank you so much, Max. Okay, I'm going to start sharing my screen. Okay, Max, can you confirm that you see it? Perfect.

All good.

Okay, I'll just wait for this to load. So firstly, thank you everyone for joining. You know, very excited to be here, to be Sharing more about you know us and our company. So very briefly, we are a location technology platform serving a range of different use cases across supply chain, transportation, logistics. I'll cover more about that. So we are a few years old as a company backed by investor like a bunch of top investors including Microsoft. Have customers pretty much spread across globally. We are Singapore headquartered, but have customers. On, I think, pretty much every continent now. Also, you know, available through a variety of platforms. So if you're on any cloud vendor, you use different TMSs or CRM systems, you can access our capabilities through this. These are just some examples of customers we work with. Okay, so now sort of moving on. So why did we start this company? Because what we saw was businesses today across a range of industries depend very heavily on location technology in one way or another. But having said that, if you use traditional platforms, there are a range of problems that come with it. So, A, they can be very expensive. It's not necessarily very easy to use, or you have to use a bunch of different vendors to essentially enable your solution. Use cases in supply chain logistics can often be very complex. Ranging from customer to customer or even warehouse to warehouse, region to region. And again, companies are not always able to customize this. For larger companies, scalability is also a big challenge. And then again, another aspect of this is to implement solutions and actually see results in practice as opposed to just a presentation. It often also needs a very close collaborative model because of all the other reasons I described as well. So these are some of the challenges that we saw and why we started this company. So our ambition here is to be sort of like the AWS for location technology. So any software modules, APIs, SDKs, tools that you need, you should ideally be able to come just to Next Billion's platform and be able to access everything that you need in one place. Transportation, supply chain logistics is, you know, probably our most, you know, core focus, although we serve other verticals as well. So yeah, we provide a range of different capabilities. Typical ones that come up, you know, that we specialize in are things like route optimization, live tracking, notifications, post-trip analytics. Also other things like let's say truck routing, maybe in a fleet telematics scenario at the end of the trip you want to do some reconciliation, You wanna see the route that your vehicle actually took. Also, you know, touching upon some of the software elements of the use case Venu talked about, you know, from a tag and track perspective, we provide a whole range of solutions to support you for whatever use case you may have. You can use these independently. You can also use them as a bundle. So we provide both options. So this is a very quick summary. Yeah, I'm just gonna go to demos now. So I have a few demos lined up. This is not fully exhaustive of the capabilities we covered, but some of the ones that I thought would be most interesting here. Okay, so the first one is I wanna showcase is a route optimization demo. So again, just like, you know, Venu mentioned as well, we are an API-first company. So what you're seeing here is sort of like a demo application. Built on top of Retool. So what you see here is a hypothetical example, one which we may all be very familiar with, where there are a set of orders to be delivered. Maybe they have a bunch of attributes like, hey, they need to be picked up from a certain location, dropped off at a different location. There may be capacity or like weight, height, volume things. There may be other aspects to it like let's say cold chain. That fresh food needs a certain type of vehicle or availability of certain resources. Maybe they need to be delivered within a certain time window, and, you know, a bunch of other constraints. Similarly, on the vehicle side, you may have other things like you have a certain number of vehicles, maybe they have different capacities, they start at a particular location such as the warehouse, they have to end maybe back at the warehouse, or maybe the driver goes home at the end of the day. There can be other things you can specify like shift times. All of these are different attributes. They're not mandatory, but this is just showcasing all the different complexities you can specify in a very simple API. So I'm just gonna, you know, run this as a sample. So I've selected, you know, a bunch of orders here. I'll select maybe a couple of vehicles. And then just press Solve. So while this is solving, so what's essentially happening on the backend is the system is running through a lot of different possibilities, trying to understand how to meet all the different constraints that we specified in the problem and how to best utilize the objective, how to minimize the cost of these deliveries. And then it essentially outputs the plan that it thinks is the most optimal for each of the 2 vehicles that it specified. So we see vehicle 1 got assigned 24 deliveries. You can see the detailed manifest here. You can also see the sequence, the total duration and time that it came up with. And you also see the exact sequence that it suggests to optimize the delivery sequence. If let's say you adjusted some of the parameters, in this case we were able to assign all the orders But maybe if, you know, the shift times were smaller or maybe the capacities were there or maybe you had a lot of unique orders, maybe you would have some unassigned stops and then you have to rerun, you know, the solver with, you know, a few different things. So it's a very highly customizable system that supports a very wide range of complexities and attributes. You can specify different types of trucks. Maybe you have hazmat material, so they have certain regulations. You may even want to run this inside of other scenarios like off-road areas. So there's a very wide range of things that we support and customizations and complexities that are possible. So this is one of our capabilities. I wanna showcase a different one, which again, it's complementary to this, it has its own standalone uses as well. So this specific one is around customization and then just showcasing how we can handle a very wide range of customizations you may come across. So I've taken a hypothetical example here in Singapore where imagine there's F1. So those of us in Singapore know how big of an annual event that is. And what I've done is I've created some hypothetical road closure here, which is that during F1 season, trucks and cars and vehicles cannot access this area. So all kinds of plans get disrupted. In a different scenario, these could be safety areas. Maybe you want your vehicles to avoid. Maybe these could be other things like certain hazmat material cannot be driven on certain roads. Could be other types of regulatory constraints and whatnot. And how this affects, you know, your system is— so now imagine, you know, at the core of everything, you wanna estimate arrival times, you wanna suggest routes, you want to track your vehicles. So if you were using a traditional mapping API, so let's call, say, TomTom, and I just picked a couple of random points around that region. So let's say, This is what I suggested. And what I'll see here is a route that TomTom or any typical location API provider would support. But the reality is it's F1, so you cannot access these roads. So instead, if I look at what we suggest, so it'll take into account the fact that these roads are closed, and therefore we need to take a very different route. Now naturally, if you were doing planning or optimization or anything around this, you know, this would have whether the routes or the planning that you did is based on the green line or the brown line, the impact would be very different. Again, we provide all of these through very simple APIs and tools like what you saw. The other aspect here is when I specify these restrictions or customizations, they take effect in real time. So if I make the change within the matter of a few minutes, you're gonna see that effect go live. And then this effect is cascaded across the entire system. So be it route optimization, tracking, predicting arrival times and whatnot, or even turn-by-turn navigation, like an app that you provide to your drivers to say, hey, you should probably follow these routes. Again, this what we see tends to be very important to be able to drive actual adoption. A common thing that we have heard from our customers is when they use a particular planning or you know optimization or any supply chain system, they don't. What on ground teams end up doing can often be very different. So so there tends to be this gap between proposed ROI versus realized ROI, and again these. Features of this kind is what we find helps bridge that gap substantially. I'll also briefly talk about, you know, another capability. Let me just reload this. So this is again just very simple live tracking analytics and reconciliation APIs. So what we provide are very simple APIs and SDKs that you can embed in any mobile device. Or like, you know, Venu was describing, tag and track, you could very well ingest real-time data, real-time location data from any other sensor as well. So again, we'll take a Singapore example where let's say we create a hypothetical asset going from point A to point B. And what I'm gonna run is a simulation here. So imagine the truck is going from this point to this point. So I'll run this simulation. So you can get a whole bunch of real-time alerts to say, okay, you know what, the truck has left the origin. Now it's sort of on the way. You can get a bunch of alerts. So maybe the driver was overspeeding here. So you can get these alerts. Maybe the driver has stopped along the way. So you get a different alert and so on and so forth, going all the way up to the final destination where you get a notification that, hey, the truck has arrived or the truck is nearby. So this is sort of the scope of this. You can take these events later and then run analysis on them as well to understand, hey, maybe there are certain hotspots your operations team need to look into. So yeah, so this is the— these are a few examples. Taking a step back, these are just some of the components that we provide. The idea is if you have any, let's say, last-mile delivery scenario or warehouse management scenario, or you want an optimization module inside of a TMS, or even, for example, you want to solve the picking optimization problem inside of a warehouse, We provide simple APIs and SDKs that can handle all of these and also fetch the data from any existing CRM or TMS system or WMS system you may have, run the results, and provide relevant data or updates back into these systems. So this is basically what I had. I'm gonna stop sharing and happy to take questions.

Thank you very much, Gaurav. Again, if you have any questions for Gaurav, go ahead, ask now in the Q&A tab, and then we'll take those questions. But let me get started with you, Gaurav. If I understand well the solutions, it's a very flexible solution, right? As you mentioned, you can easily pull data from whether it's a TMS or any other tools. And also, how do your clients actually use it today in their operation in Asia?

Sure. So a common scenario is just route optimization. So let's say you need to make deliveries. You have a warehouse. Imagine you're a grocery delivery company. Now you have a bunch— your customers have picked certain time windows for deliveries. You have orders coming in and maybe the operations team have Excel sheets, who knows? And then they have a bunch of constraints. They have different planning sequences. So they may say, hey, at the end of today, I want to plan deliveries for tomorrow. So I start planning at 5 PM. But then more orders have come from 5:00 PM to 10:00 PM. And just before the truck leaves at 6:00 AM tomorrow morning, I want to incorporate those fresh orders as well. But shit, 5:00 AM, I learned that one of my drivers has actually— is not going to show up, so I need to readjust these. So we provide these APIs where you can specify all of these things and you get results in real time. So this is the idea. So in this hypothetical example, we provide a tool where they can upload a CSV as well, which again on the backend just calls the same API. They can get the results and use that. If they have developers on their team, they can very well just integrate it directly with any internal tools that they have. So that's a hypothetical example.

Okay, so are TMS not able to do such thing today?

Yeah, so a bunch of— so there are 2 parts. So TMS companies, or TMS providers, especially the larger ones, may have these modules. Many TMS providers, they don't. Even when these modules are present, what we have seen is in reality, the delivery optimizations or operational plans that companies have can often be extremely different or far more complex than what our, you know, off-the-shelf solution provides. And therefore, what ends up happening is even though companies purchase these solutions, they don't— the actual teams are like, okay, you know what, I don't think this fits what I need to do at this warehouse or in this region. So I'm going to stick to my original manual planning. Or at best, I might do— I look at this and I'm like, yeah, sure, let me see if I can run my own thing because I've been doing this for 5 years. I know how this works better.

Okay, we have a question from Akhilesh who's asking, are you using any solvers or heuristics for the route optimization? And he's adding, and how does it cost to a sponsor?

Yeah, so yeah, I'll answer both questions. So yes, we have written our in-house solvers. We also own the underlying maps, so map data, mapping APIs. We also have other companion APIs like territory management, geofences, zone management, etc. So we've written all this. This is all proprietary software. And yeah, so this can be plugged and played into different systems, handle different scenarios and customizations. In terms of cost, we offer 2— the typical thing is API, so it's usage-based pricing, but we are very flexible to fit different business scenarios. So I can give 2 examples. Taking the same grocery delivery examples, we can price on per asset. We can also price on per order. So which is just to say, okay, you know what, we'll charge you a fixed amount per order that you deliver. And within that, however you want to use this, you want to run different simulations, you want to run different simulations every week, all that is covered just for a fixed cost per order or per asset.

So obviously the quality of a map is very important. Is there any countries, for example, in Asia where you struggle to get accurate Map information?

So it's a great point. Usually, again, we work with third-party partners who pretty much solve this at a global scale. But if you have to ask me, like, what's— so once you get to very small countries, that's when sometimes it can be a challenge. So if you ask me, how is our map data in Myanmar, could sometimes be a problem. If I look at general India, South Asia, Southeast Asia, Vietnam, Malaysia, Singapore, Indonesia, it's all pretty good.

Okay, we have another question from Akhilesh asking what sort of volumes can be handled in a single call? How much time does it take for a call with 1,000 shipments with 6 to 8 constraints? And maybe you could explain more, it's a bit technical here.

Yeah, yeah, sure, absolutely. Yeah, so thanks, Akhilesh. So basically the question Okay, a common problem with these systems is that while they can work to handle 10 orders or 20 orders, as the number of orders grows, if you're a larger company or doing lots of deliveries, very often it can become very challenging or really slow, and or sometimes they cannot handle it at all. And what that ends up meaning in practice is you cannot do deliveries the way you want to or run operational playbooks the way you might want to. So, Akhilesh, to your question, scalability is one of our biggest advantages and strengths. So yeah, we've done anything up to like 5,000 shipments in a single API call. Obviously, in terms of the time it takes, it scales with It grows with the number of shipments. I don't have the 1,000 number off the top of my head, but I remember the 205,000 number. So extrapolating to 1,000, I think it'll be around 10 to 20 minutes.

Okay. So there's a lot of talks about AI these days, ChatGPT, AutoGPT, and all those different tools. How does your solution integrate that? And, you know, I've been looking again at you your demo, I was thinking, okay, could you just ask, you know, through chat the system to, you know, optimize certain things for you and then it can do it automatically for you? And could also your platform actually by itself, you know, suggest some kind of optimization?

Yes, absolutely. A great question. So that's another one of our what we call modules, which is, you know, essentially AI-powered solutions. So another thing, and it all goes back to driving adoption of these solutions by actual teams. What we have seen is if we can leverage customers' historical operational data, then we can make solutions far more compelling. So we have machine learning and AI capabilities where if you have been doing deliveries or you have telematics data from your devices and vehicles, and how your teams have been doing these things in the past, our machine learning capabilities can learn from that and automatically suggest that, hey, maybe, you know, you've been suggesting this route to your drivers, but your drivers typically, you know, in these times follow these routes. Or you think you can do 10 deliveries in one vehicle in a day. What we have found is in reality it's only 9. vehicles, you know, 9 deliveries that can be done. So, so yeah, these are all different capabilities that we also provide.

That's interesting. Okay, that's great. So in terms of coverage with your solution, you're covering the entire Asia region, right?

Yes.

Are you also supporting your clients in the US or Europe?

So we have customers globally, so South America, North America, Europe, Asia, Australia, New Zealand, and parts of Africa. We don't have complete coverage in all African countries, but yeah, pretty much the entire globe.

And who are your typical customers for this kind of solution?

So we have 3, uh, so focusing on, let's say, transportation supply chain. So we work in 2 ways. One is, let's say, let's say tech companies such as, you know, I imagine a bunch of our audience. So if you're a tech company, you have the capability to use APIs, SDKs yourself. That's one customer category. A second category is we often partner with ISVs, etc., to sell to very large traditional companies. The third category is very big enterprises with in-house IT teams can also integrate this directly.

Okay, thank you very much, Gaurav. Oh, there's another question. I was about to wrap up. Akhilesh is adding also. So in the past, there's been issues. I've seen issues with passing coordinates of a customer and an input for optimization. How do you address the same in B2C scenarios when addresses are not accurate or difficult to triangulate? Good question.

Yeah, it's a good question. Unfortunately, we don't have the perfect solution for this. We offer some capabilities to help reduce the impact of bad addresses and essentially think of it as a percentage improvement. We have a product that we call Multi-Geocode. that allows our customers to manage their own intelligence. So let's say, Akhilesh, if your teams have been doing deliveries in the past, you know, in certain neighborhoods or for certain addresses, again, from our machine learning system, we can then understand based on your historical deliveries that, hey, this address most likely means this location. So that's one scenario. In B2B scenarios, like let's say a Procter& Gamble doing deliveries to a fixed set of retailers or distributors, that is easier because it's a fixed, you know, set of distributors or retailers and we can triangulate that more accurately. But B2C is— there is no good solution except over time with more of your data, we can help reduce the impact of bad addresses.

All right. Thank you very much, Gaurav, again for sharing, you know, what you guys do at Next Billion AI. Very interesting solution. I want to thank you again, Venu, from Tag&Track, also to join us today. I want to thank both founders for taking part in this Demo Day. We will do this again. We hope that you enjoyed it today. And again, we will follow up with recording. And if you have again any questions, feel free to contact either Gaurav or Venu directly or their respective representative to get more information about the solution. Thank you guys, and we'll look forward again to see you next time.

Thank you so much.