Today, on How AI Happens, we are joined by the Chief Digital Officer at Allied Digital, Utpal Chakraborty, to talk all things AI at Allied Digital. You’ll hear about Utpal’s AI background, how he defines Allied Digital’s mission, and what Smart Cities are and how the company captures data to achieve them, as well as why AI learning is the right approach for Smart Cities.
Today, on How AI Happens, we are joined by the Chief Digital Officer at Allied Digital, Utpal Chakraborty, to talk all things AI at Allied Digital. You’ll hear about Utpal’s AI background, how he defines Allied Digital’s mission, and what Smart Cities are and how the company captures data to achieve them, as well as why AI learning is the right approach for Smart Cities. We also discuss what success looks like to Utpal and the importance of designing something seamless for the end-user. To find out why customer success is Allied Digital’s success, tune in today!
Key Points From This Episode:
Tweetables:
“I looked at how we can move this [Smart City] tool ahead and that’s where the AI machine learning came into the picture.” — @utpal_bob [0:11:11]
“[Allied Digital] wants to bring that wow factor into each and every service product and solution that we provide to our customers and, in turn, that they provide to the industry.” — @utpal_bob [0:16:27]
Links Mentioned in Today’s Episode:
EPISODE 46
[00:00:03] RS: Welcome to How AI Happens. A podcast where experts explain their work at the cutting-edge of artificial intelligence. You'll hear from AI researchers, data scientists, and machine learning engineers as they get technical about the most exciting developments in their field and the challenges they're facing along the way. I'm your host, Rob Stevenson. We're about to learn How AI Happens.
[00:00:31] RS: Here with me today on How AI Happens, is the former head of AI over at Yes Bank, a litany of titles in the AI space. Currently, he serves as the Chief Digital Officer over at Allied Digital Services, with Utpal Chakraborty. Utpal, welcome to the podcast. How are you today?
[00:00:47] UC: Hello, Rob. I'm good. Thank you so much. How are you?
[00:00:49] RS: I'm really, really great. I'm excited to be meeting with you. There's a lot of different directions that we can take this conversation, because you, in addition to all of the work you're doing at Allied, you have a lot of just experience in the AI space. I know, I'm just a kid in the candy store, because where is this going to go? But excited to learn from you today. Would you mind for the folks at home sharing a little bit about your background and how you wound up in your current role at Allied?
[00:01:12] UC: Yes, thank you. Thank you, Rob. My journey with artificial intelligence started around 15 years back, when I was thrown into a project. At that point of time, I was not much aware of the different technologies that come under artificial intelligence. I took that as a challenge. Then a lot of new things, mainly linguistic area, NLP, and then machine learning and deep learning. We successfully delivered that particular project. Subsequently, we got to multiple other projects. That's how I started learning. I started doing some research on artificial intelligence and then moved into different career ladders, like heading the artificial intelligence and the analytics vertical in years back, so that is a banking domain.
A lot of experience there was you have got a huge amount of data. Actually, you can play around with data and you can actually do a lot of things with AI. Then moved into Allied Digital. In Allied digital also, there are huge amount of data, because we have got Smart Cities, where you've got data from different sources. Also, you've got a FinTech article in Allied Digital, wherein we do R&D in FinTech, we build FinTech solutions. We have developed a program called FinoAlled, which is a conversational banking platform. So this is how the journey is.
[00:02:34] RS: I see. Would you share maybe, how would you characterize Allied Digital's mission?
[00:02:40] UC: Allied Digital is a customer-centric organization, wherein the customer is at the center stage and all other products, services, whatever we are offering are all aligned towards the success of the customer. That's how I can define it. That is there in our DNA. The news and technologies that we implement today are on the cloud. They are AI solutions and services. Of course, we have got Smart City as I told you, and also, traditionally, we are a Master System Integrator, okay. That is also our business. But there, also, all the particles you can look at, okay.
What we are trying to do is, we are trying to bring that automation and also, that cognitive element into the entire workflow, or the process. That is our idea there. If that load-like artificial intelligence, please environmental role. If we talk about Smart Cities, Smart Cities, we have got command and control center. This command and control center if you look at today, you cannot categorize it as a very smart command and control center. The cognitive elements are there, but it is not ahead of us. There is a huge scope to bring in AI into that, as well as bring in lot of optimization in that particular area.
Similarly, when we talk about even the infrastructure, there are areas where you can do infrastructure analytics. You can have many predictive elements into your infrastructure parts, so that before something goes wrong, you can actually analyze through your AI and you come to know that, hey, there is a possibility of this kind of a problem and this particular server on this particular business. This is going to be a game-changer. Although in the industry, I hear there are some works are going on, but I don't think it is still an interventional way. Where we want to take a fast-mover advantage. We want to bring in solutions, where basically, we can do all predictive analytics in that area also.
In the FinTech area, of course, because in FinTech there are huge opportunities where you can implement automation, you can actually bring in things which will empower the banks, especially the small and mid-size bank, to actually render their products and services to the end customer in a similar kind of a manner. Also, even to the lowest grade of a customer society, you can see, because many of the unbanked people, if you look at today, especially countries like India and many Asian countries. So how to take your banking services and products to the end customer who are in the lower grades, right?
Our platform actually provides that kind of facility, where you would put a minimum investment, banks can transform their products and services into physical products or services. There is and probably those are physical services, and they can deliver those products and services to their end customers, wherever they are residing, probably the remotest villages somewhere in India, or into urban areas, [inaudible 00:05:52] areas. These are some initiatives that we are bringing a lot in there.
[00:05:58] RS: Can you share what you mean, when you say Smart Cities? I want to get into the Allied opportunity with crunching this data from Smart Cities, but I guess, it's probably useful to define it first.
[00:06:09] UC: Yes. So, Smart Cities and we talk about today. So you can say, it is an intelligent city. How to bring an intelligent city. By the way, we are pioneers in Smart City in India, okay. So many of the Smart Cities we have built in India. I think, there are two parts to it. One is definitely the infrastructure part, where actually, you put all your infrastructure in the Smart City, okay. Then bring all those information into your command and control center, wherein you can actually control and deliver commands to the parts whenever required.
Apart also, this is a traditional way of doing that and probably, in many countries, this has already happened. But now, we want a step ahead, okay? But how to bring that intelligence into a Smart City? So if you are providing some kind of a citizen-centric app, right? Where a citizen looking at the app, a citizen can understand if they are traveling something, what is the traffic condition? Is there any incident that happened at that particular place? There are many other things. Those things can only be done once you pump in all your data from different sources, okay. It could be your traffic management system. It could be your different CCT can browse it. It could be some other sources. You actually put it into a data lake and management and build some data marks and you run analytics on top of that, okay, and bring out the intelligence and provide it on demand to the citizens. That is one part.
Second part is the optimization. Optimization is something which has been not looked into in a very serious manner in many of the sparsity projects that I know, even across the globe. I can see as an AI explorer, I can see there is a huge opportunity for doing optimization of resources. It could be your electric resources. It could be water resources. It could be anything or anything, right? So that is also possible, by implementing machine learning into different parts of your Smart City solutions.
[00:08:11] RS: Would it be necessary to distribute sensors all over the Smart City? Or would you be able to synthesize from existing closed-circuit television, satellite, etc., or how would you manage to capture this data, I guess, is my question?
[00:08:26] UC: Yeah. It is both that ways. I think that optimization is very, very important like where from actually you need to take the data. Of course, some information that we do need to have this, so for example the garbage management system. On your garbage management system, the garbage being said, all those things, you cannot rely on satellite data, right? I mean, satellite data is going to give you nothing. They’re a bit sensitive. I'm just giving you an example. A flat situation, right. They're probably, I mean, it's the mixture of the satellite information and the information that you have with your IoT sensors that are embedded across the Smart City.
It is like both the things and on top of that, run machine learning and come out with the right kind of assessment of the situation. Then you inform the administration and the citizens accordingly, so that they can take care of themselves accordingly.
[00:09:20] RS: Got it. Looking back at just Allied’s role in developing Smart Cities, or even in rolling out these FinTech services for you at a high level, what made it clear that AI machine learning was the right approach for these services?
[00:09:36] UC: Yes. Traditionally, many of the things are already happening. It is like for example, in the FinTech, FinTech is a new division and in this new division, I mean, AI machine learning is something which is required, because I think this is conversational AI, conversational banking. Definitely, AI element is very, very important. Apart from that, we picked up on other initiatives. What I found is, things are already happening and we are delivering our services to a great extent up to the satisfaction of the authorities. But how to go beyond that? That’s where I think you can add extra value to your services, or products for that matter. That's where this cognitive technology comes into picture.
For example, I'll give you another example also. You've got a product in service management, okay? It's basically, service management system. It's a product. Now the service, the name of the product is Adidas. This product, we are already 100-plus customers, many big customers are there. Customers are happy, because this is slightly flexible tool in terms of creating tickets, and then the entire lifecycle of the tickets and creating dashboards, looking into that. Managing those things is very, very – even in the command and control center also, it can be integrated.
If any incidence comes, then you have coding workflow, right. It goes to this particular team, if there is all done, it goes to Allied team. Then that particular incident is closed. But where I looked at it, was this the brilliant tool, but how we can take this tool one step ahead and that’s where the AI machine learning came into picture. I thought, okay, fine. Now, I mean, how can I give voice commands, so that in Internet language, if I have to a system, system should understand and system should do that. That is the kind of thing that is happening.
In FinoAllied FinTech platform, it is actually completely powered by AI. We consciously took that decision that let's use AI as much as possible, because futuristically, probably, today, some of the AI components we can eliminate, because probably, a rule engine can do that. But if you just take a – if you look at, we’ll have to use AI, because probably right now, we don't have that much of data, but tomorrow, definitely, we are going to get data, and our models will be much more, more [inaudible 00:12:00]. Extensively, machine learning models in AI has been used across the FinoAllied platform, so that we can make this platform very, very seamless, very, very plug and play kind of a thing. Very, very inclusive in terms of giving any analytics.
For example, I will tell you, so many platforms are available. Wherein, if it is implemented into a bank, bank gets all kinds of analytics where they are required. How many food deposits have been to – okay, these fixed deposits are coming from which segment of the customer, and how many transactions are happening out of that, how many formula and transactions are happening. Then those kinds of information frequency, in it as well. Those analytics, everybody's giving. Now, what we did is basically, we are giving customer analytics.
Now, as a bank customer, you can have all analytics, just you're giving a voice command in your binocular, so suit is more completely multilingual. You can keep that analytics in front of your eyes. What is my utility expenses in last quarter? Okay, so now in utility, there are many things comes. Your electricity bill, your water bill, your Internet bill, your any other things, this all comes down to the utility. My machine learning model understands, what is utility? What it is doing?
If I say, what is my trend in spend in entertainment? Automatically understands. Like, my six months last six months, what is the trend of my expense in entertainment items, like movies, theaters, this, that, all those things. It will also show you the trend goes in this way, then probably for next three months, this is going to be the trend. This is going to be a game changer and it is already in the banking industry, people are liking a lot. This is going to be the game changer.
[00:13:45] RS: Yeah, it's interesting, because you're deploying these FinTech products. Also, you think a lot about the opportunity of just optimizing every process. When you look at complicated organism, like a city, for example. I guess, I keep coming back to this. It sounds like Allied is trying to boil the ocean in a way. When you're at a company that size, you can, because you have so many resources at your disposal.
For your approach here, is it just efficiency? Is it optimization? What does success look like and what is the end goal for you? When you're looking at how do we help people understand trends about their personal spending, but also, how do we help garbage trucks have more efficiency? What is the ultimate throughline connecting all of these together?
[00:14:30] UC: Yeah. Ultimately, we categorize ourselves as services industry. We are a software services. Here, a customer success is our success in that way. That customer could be a Fortune 500, it could be a bank, it could be anything. Wherever either we are giving our services, or our product is running in their environment currently. I think, all those things, some of the things are already given. As I say, the automation. Yes, of course, if we don't do automation, you will not be able to do savings. Every customer, today they are looking for savings.
Second thing, if you don't do automation, you will not be able to bring their facility into their processes, which is very, very important today. Your backlog should be strong and your backbone processes should be strong enough to support your font-end processes, so that you can actually make the entire system design for your customer. I think, the most important thing, which the entire world is basically after today, use specifically the customer experience. Not customer experience as a whole and in this word is a very prophetic word. It has many meanings, and it has been used interchangeably here and there.
What I feel is, if you want to give a customer experience, there are two parts to it. One is the utility part. Okay, so what you're very intuitive you're providing to the customers? Whether that customer is getting that utility in a seamless manner. Okay, so that is one part. Second thing, and which is most important thing is the wow factor. Probably 10 other people are providing that services seamlessly. Okay, but that wow factor is not there. How you can bring their wow factor in that. That is only possible if you use a technology like artificial intelligence and machine learning. That's what our aim is in Allied Digital. We want to bring that wow factor into each and every service product, solution that we provide to our customer. In turn, they provide to the industry.
[00:16:39] RS: Yeah. You touched on something interesting here, which is that it's not simply enough to develop a fantastic technology, right? Certainly not in the private sector, because people have to use it, people have to understand it. This design thinking approach, I've heard it called, like thinking with the end-user in mind. It sounds more and more like that's everyone's job. It's not enough to be a fantastic data scientist, a machine learning engineer. Every person in the process of developing this technology has to be thinking with the end-user in mind. Is that how you think about it at Allied Digital? Obviously, because you’re customer-centric. I guess, my question is, whose job is it to be putting themselves in the shoes of the end-user and designing something that is seamless for them?
[00:17:23] UC: Absolutely. I think, it is the job of everyone. Although, you have got a CX team, who are expanding to the customer experience side of it. But your developer should also understand it well. If your developer is not understanding – I think, it is the entire journey of your product, or solution development and the people who are involved in to different stages. It could be your developer, it could be your tester, it could be your experience people. Everyone should be aligned in that way that this is the experience that we are going to provide to our customer.
I mean, when we talk about customer experience, there should probably an ambiguity. What kind of experience? You should be able to define it. You should also be able to quantify it for a particular thing. If you cannot put this in a paper, that for this particular solution, these are the features and these are the extra factors that I’m giving it to the customer. If you cannot quantify it, I think in that case, I think you will not be able to provide that customer experience.
I think, many, many, many times I have seen, people are confused when we talk about customer experience. There are many things that comes under that. I think, the best way to deliver it, basically define it, that these are the things that I’m going to define in this particular manner. Whether an up post, and also see whether it is what – CSAT is very, very important. Now, if you look at our FinoAllied product, it is a banking transactional system. For each and every transaction, if somebody sends some money, if somebody sends some PSA utility bill, or somebody books a fixed deposit, or a current deposit, for each and every transaction, what we take, we take the customer with them, okay, in a soft manner.
Because today if you look around, if you ask anybody to give a feedback, okay, so they will not be able to give a hard feedback. It is very difficult. You make it a soft feedback by putting some emojis, okay, or some thumbs up, thumbs down, or some different – there are different mechanisms there. One to 10 scale, you select something, and you get that and ran analytics on that and see every transaction, how people are liking, disliking or is it okay? Is it a wow? That's very important, I think.
[00:19:36] RS: Yeah, of course. Utpal, before I let you go, I wanted to just ask you about the AI space writ large, whether it's at Allied, or outside Allied, could you share something in the space that has you truly excited, maybe a possible use case, or just development in AI technology that you think is going to be truly disruptive?
[00:19:55] UC: Yes. Today, if you look at almost every industry, AI applications and AI implementation is improving. People have already understood. At least, if I talk about in the Indian context also. A lot of AI implementations are happening across different industries. Having said that, as I'm coming from a finance, banking and finance industry, there is a huge opportunity there. Already, lots of things are happening there. Starting from your customer-facing applications, really, your back-end applications. There are a lot of AI use cases.
Even the embedded AI, wherein basically, I can embed my AI elements into the very, very core banking applications, like core banking applications, LMS, loan management systems, my credit card systems and all other systems. That embedded AI is also happening. Even FinTechs. We are basically tying up with banks, and who are dealing with mainly the front-end part of the banking. Now, they are also finding opportunities to implement AI into the other areas, which are not front-end, or customer-facing, which are the core business of the bank. Bank of halogenation. When I talk for the bank, insurance and all other financial issues. There is a huge opportunity here.
Apart from that is healthcare and medicine, another area where a lot of things are happening. I know some of the startups, or some of the companies, they are also coming up, and they came to me for different kinds of advices on their applications. There are also a lot of good things happening. I'm pretty excited on the final side. Also, this IoT adaptation, which are like Smart City. This is going to be advantage. Also, to tell you now, web 3.0 is the meta war system. Here, once web 3.0 is implemented in a democratic manner, at that point of time. AI is given in web 3.0. It’s the integral part of the 3.0.
Now, we will see any application you develop, AI element is going to be there. I think, it's a very exciting time for artificial intelligence, if you talk about. Of course, other subsequent categories are also there, because when we talk about web 3.0, AR, VR, blockchain and all other contributing technologies are there. AI definitely is there and it is going to play a massive role into the entire state.
[00:22:33] RS: Utpal, this has been fantastic chatting with you today. Thank you so much for taking the time and meeting with me. I've loved learning from you.
[00:22:38] UC: Thank you so much, Rob. It was really great talking to you.
[00:22:42] RS: How AI Happens is brought to you by Sama. Sama provides accurate data for ambitious AI, specializing in image, video and sensor data annotation and validation for machine learning algorithms in industries such as transportation, retail, e-commerce, media MedTech, robotics, and agriculture. For more information, head to sama.com.