HCL Technologies Limited (HCLTECH) Earnings Call Transcript & Summary

Good day, everyone. My name is Surendra Goyal, and I head Research for Citi India and cover IT services as well. Welcome to the webinar on decoding the impact of generative AI on IT services, thoughts from HCL Tech. Before I start the session, let me highlight, investment in securities market are subject to market risks, read all the related documents carefully before investing. Since the announcement of ChatGPT and other Gen AI initiatives, there has been a lot of discussion around generative AI and its potential for IT services business. Secondly, any disruption it can cause to existing businesses? In fact, this remains one of the most discussed topics in our discussions on Indian IT with investors in the last few months. To get a better understanding of the impact, we are hosting Kalyan Kumar, or KK as he is commonly known, Global Chief Technology Officer and Head Ecosystems for HCL Tech for a deep dive on the topic. KK is the global CTO and Head of Ecosystems for HCL Tech overseeing all product and technology strategy, emerging technology incubation and the overall Cloud offering, CloudSMART. He leads the HCL Tech ecosystem units spanning Cloud, Tech OEMs, telcos and the start-up ecosystem. KK is also the Chief Product Officer for HCL Software. He is responsible for all portfolio and product groups driving central engineering, cloud and SaaS platforms, IT, security and compliance for HCL Software. Over the past 2 decades at HCL Tech, KK has played various roles across solutions, business development, alliance management practice, COE and delivery, starting at HCL Tech subsidiary, HCL Comnet in the year 2000. The format we'll have here today is a presentation and initial comments by KK, followed by a moderated Q&A. I'll be asking the questions. If any of you have questions, please e-mail to my e-mail id, [email protected]. For people attending on Citi Velocity or the webcast, you could type in your questions as well, and I will take the questions from there. Welcome, KK, and thanks for doing this for us. Let me just hand over the session to you now for the presentation.

Thank you, Surendra. Good morning, good afternoon, good evening, all people in all parts of the world. In [indiscernible] today, I'm in India. But generally, my home base is in the U.K. It's great, and thanks Surendra, for this opportunity. I have done at least a dozen such conversations now with customers, clients, Boards, with various differences [indiscernible] and every time, it's such a diverse audience, I always trying to learn something. It's like every time we try to learn something in our conversation. A few weeks back, I learned something new. So I'm going to just share my perspective. Perspectives are continuously evolving because the tech stack itself is evolving. But it's interesting to see the potential and what new opportunities which are there for technology services industries and also as a software product, [ IAST ], we're also seeing some very interesting perspectives. I'll share both sides of this, Surendra. So maybe -- I'll talk about the next 35 minutes to give you some perspectives on this. So this is what we're going to do. We're going to just try to demystify generative AI. And then really talk about different value streams of how we apply this across a different service capabilities which exist in the IT services and engineering services industry. And then we look at what are we doing at HCL Tech Services and then what are we doing at HCL Services software around this, and then we can pick on from there. So I think the first thing, which in a lot of my conversations, the media and the market somehow has hiked generative AI as the super set of AI and somehow people think it's something just a broad thing, which is going to be over encompassing AI. But if you really go and I always suggest that read Chapter 1 of Professor Stuart Russell's book because he is, I would say, currently the most living legend on defining some of the base fundamentals of what AI is. So there's a large discipline of AI, correct, it's not new. Within that, there is a discipline like you have computer vision, robotics, a lot of other pieces. Within that, there is a discipline just called machine learning, when you have supervise, non-supervise. Within machine learning, you have a subset discipline, which is deep learning. Actually, generative AI is an intersection or it's basically at the confluence of what's -- and large language models, it's basically the foundational underpinning of [ G AI ]. It's basically, I would say, at the confluence of deep learning and natural language processing, ability to combine both of them to learn using the deep learning methods, but the new process language, understand language and generate language. And then there is a lot of other use [indiscernible]. So it's primarily a subset. It's very, very important. So basically, you can take image, pictures, codes, script, text or document. I didn't use the word video because if you go back and see, video is nothing but images processed within a second. Finally, when you se video, there is a 30-frame, 60-frame space. Images are fun. It's a fundamental building block. And you use this source of data to train the model or the language model. And that's what deep learning comes. It learns the model at very deep -- this thing -- understanding. It's basically learning just like [indiscernible]. And then you basically use prompting and query to basically improve what the language model has learned. If we ask questions, you get responses, you tune and it continue to do that. And then it generates output and this output could be a lot of different output; codes, test cases document. So this whole -- chatGPT, what really got -- brought in, was really -- and I would say, the first major public consumer implementation. So what we are seeing today, whether it's chatGPT or Bard or Bing chat or now Google has chat plug-ins, which is available similarly with their Bard in their search and they're releasing out in the same way. It's basically the consumer-facing world, correct, which is all out in the consumer. All of us can use it. Those language models are trained on public data. So if we see chatGPT many times and ask the question, just I have been trained till 2021, correct? I can't answer anything before or after that. It gives you that very standard disclaimer, like Surendra put a disclaimer at the start of the session, correct? So it gives you that, but then what's happening is as you see evolution, like what Bard is doing. Bard is actually generating on top of Google search also. So it can search context, data -- real data, and then it's able to generate on top of that. It's able to find -- and it's continuously evolving. Look at what Microsoft has done with Bing Chat in the consumer side, sticking in the chatGPT capability. But that all if you use chatGPT plus, which is the subscription, the paid subscription version, you can use different language models that can turn on plug-ins and you can see different plug-ins coming in. So that whole thing is evolving in the consumer space. But I think as collectively from our perspective, we need to start to realize it, okay, consumer space is going to adopt it at scale. A consumer product and tech companies, which are index space will embed this technology. That's the lot of the work, which we see as engineering services from our standpoint. We are working with a lot of those companies to see how they are adopting and implementing the technology. You want to see these interfaces come up with your consumer electronic devices and your various different things very soon. But then there's an enterprise view of that, and we touch upon that so because most of the IT services firms have a big play in the enterprise customers, correct? That's where the adoption is going to happen. When we look at -- you look at adoption on 2 lenses, correct, the consumer lens and the enterprise lens. Now, let's pick some 3, 4 areas. So this is what it is -- what I thought of doing was to pick those 4 common -- I think I picked up the nomenclature, which Surendra came out of your report, like you call BPO, IMS, App Dev and we also added a section on engineering, correct, systems engineering, very specifically. And typically, a BPO, your front office, your middle office work and your back office work. Where -- if you really look at it and just start to look at the value stream of the BPO, there are a lot of activities, which we do, correct? A human does a lot of activity. So many a times, again, we miss -- sometimes misinterpret the context of, it's a human machine partnership. It's like, Microsoft call it, Copilot. Google calls it, Duet. We call it the human in the loop or the human machine partnership. So it's -- basically it's human and the machine, correct, it's not human versus the machine. And the whole process is that within that -- in the BPO segment, you do process analysis, execution, process improvement and you deliver value to the customer. There are areas where you can apply Generative AI, like search and retrieval, anytime someone ask for a query, search and retrieval, can you intent extraction from a document, correct? Like you can use OCR, but also you can read the document and extract the intent. There's a lot of things. Because anything which you do with a language, you're able to apply it, correct. You could generate standard operating procedures like, if there's a complaint or things like that, how do you handle this. You can generate a lot of text. You can summarize knowledge. You can translate languages. You can ask for first level conversations, you don't need. You can use this to argument some off-peak loads and say you can translate language, help in process engineering. You can see things which you have to read a lot of documentation. There is a lot of compliance type work, they're reading a lot of information, trying to extract information, summarize, look at things. So anything which a language model can do, there is an applicability in the value stream in the BPO space, correct. So like if you look at that. So I think the input could be case, issues, request, anything just coming in documents, all the things you talked about before, correct. So that's one in a very simplistic way. There is applicability. [indiscernible] to the same thing. If you could go back, Surendra -- RPA, remember about 10 years back, there was a huge [indiscernible] around RPA and then RPA became CPA, correct, oh, I'm going to insert some machine learning, some cognitive. You look at the adoption of RPA, it's starting to a good level, it's taken a good amount of time because when you get into an enterprise rollout, connecting to various different systems, tying them together, integrating that data, cleaning up data. I'm going to talk about some of those prerequisites or areas which you need to do that. BPO is one area where you see this is a typical impact. Second area, there's a lot of talk about it can generate code, correct. This is a big, big debate of the code generation. You see most of the public implementation of the Codex model or the data Copilot or the Google code gen or generative app builder. These products have a lot of affinity to modern languages because we've been trained a lot of them into the modern programming languages and they can generate a lot of scaffolding code. We can generate a lot of code, which is basically nonfunctional. And if you train them with a lot of functional know-how data, it can generate a good amount of code, but you still need to validate that code, correct? So again, this is -- somehow people think you can write code and the code can be taken and deployed into production, correct? It needs to go through -- so when you do an application development process, correct, you have requirements, features and bugs. You do analysis, process mining, understanding, experiment pilot, you model and develop. There's a lot of activities which you do. I think the lot of -- I have a slide coming up later, which gives you some range of how the coefficients is, but there is a lot of areas around generating test cases, generating scaffolding code. It can also help you do some level of code generation for programming languages which you might -- like the developer might be trained on two, but I think if there is a variance of that programming language that he or she wants to generate a code. Maybe I'm a NodeJs developer, but I want to generate some [indiscernible]. You can do some level of code translation. It's very proficient at some languages like Python, which is now, I would call more like -- it's like a more citizen-ish development language because I think most of our excel sheet users in -- even in your industry are now all becoming writing -- sort of writing macros, now they're loving to use Python to manipulate data, correct? So I think it can generate some good level of code. It can ingest a lot of documentation like application support kind of this thing. So when you're troubleshooting a customer, you can -- rather than you searching up a document, you can actually quickly look up something in supporting this thing. But in a dev, there is some significant touch point, which is there. But in pockets, correct, again, when you do the whole ADM life cycle, there are a lot of things which you do. Business requirements, and now it's agile, what's happening. You do so much of iteration, correct, going back to the customer, working on prototype. So you can do some UI generation, correct, because you can do image generation. So you can generate UI, you can generate some UX flows. You can do a lot of work. I think it can assist -- so the way one has to start thinking is, how do I pair it up and use it in a far better way, correct, rather than thinking it's and/or -- I always say, it's the -- and how do you partner and do that. So there's a lot of areas within application development and support where there is a significant applicability, which is there for this. And again, when you contextualize -- now, when you do app modernization, it has to be trained on a lot of traditional languages, correct. Like if you see COBOL, it's an interesting language. People think there is still a lot of -- I think we're still talking about Gen AI, but there's still gazillions of mainframes running around there, running COBOL code and there's an opportunity that as COBOL programmers are not available, would you create a good COBOL Copilot, which can help you generate code, can help you modernize those code. Can you use this to convert from COBOL to Java? The interesting question is that that's been a mythical challenge because COBOL is such an interest language it's got business context and nonfunctional weaved into one programming construct and how do you extract. So there's still -- there's an opportunity. It's always -- just like an automation, you keep improving. There's always an opportunity, correct, which exists for a long period of time. Let's look at infrastructure and operations. Again, in that, if you look at OPs -- see, there are a lot of infrastructure work like projects and all. We can't do this with that. Kind of same thing is a lot of project build, like you implement packaged applications in application development. Yes, you can do some documentation, initial work, but we need to implement package applications in lot of -- now if the package application vendor is embedding some generative AI helping them implement, that's where they will be able to speed up the implementation. But you can't use the Gen AI from our side and say I will implement a packaged app, correct, an ERP or CRM or this thing because you are commerce system. You need to have context of plug-ins to be able to generate that, but things where are possibility, there is applicability. [ INO space ], this has always been the favorite child. Every time people say automation, first time -- if we go back 10 years back, like an RPA example in BPO is the cloud. Oh, will cloud change the IMS [indiscernible] realize in the enterprise adoption, it's actually the migration and modernization and some cloud is becoming a very, very important vehicle to help customers move that this thing. If you look at [ INO space ], there are areas, especially in operations, correct, where on top of what you do with automation, with AI and ML, you can apply a lot of generative AI capability around analysis, extraction, RCA, doing RCA, intent extraction, script generation, a lot of automation script generation, SOP generation, you can assist in threat modeling in case of cyber, it can improve a lot of employee experience by giving them more conversational bot which can talk to various different systems, knowledge summarization, language translation. So we were at the [ INO space ]. So again, the same logic of you talked about cloud, correct, on the cloud moment of adoption, it's going to -- and the way we got automation embedded into this. There's a lot of areas where you can apply this. Again, you have to contextualize and again, when you start to use this in the customer environment, you need to, again, make it enterprise-oriented. We'll talk about how we do that, correct? There is also some play in the systems and product engineering. Areas, it can assist in areas like factory automation, VLSI design, can generate [indiscernible] circuit diagrams, can help in UX design content generation, can support in code testing and optimization, can support some of the Level 1 product support kind of activity. So even in areas where -- because as softwarization is happening in most of these embedded systems and connected assets, you have to realize that the total number of code which has been ever written until now, you're going to have 5x, 10x of more code being written as more and more of these devices get softwarized, so I think there is a huge potential. And I think humanly whether you've applied and got software product engineering, every company wants to become a software company, correct, they are reorienting themselves to build more and more software digital products. Silicon and mechatronics, there are areas with as and when those software control plane versus silicon. There's a whole evolution happening around how do you control silicon using software and how do we embed more software in silicon, and some cases, full software and a lot of silicon into the cloud. You see a lot of similar plays happening [indiscernible]. So I think there is applicability a bit there. So what we did was -- the way I really look at this is we have to see this in 2 perspectives. If you -- all the different areas we talked about, there is a lot of AI, ML and automation applicability, which exists today in various activities that you perform in these services. Generative AI has potential to extract and pull that to a certain -- it give you more efficiency. And then if you see this quarter, and again, if you generally go and read reports, there is no sanctity in terms of -- if you take a piece of work within the larger context, it's just like test case generation, then they'll say, oh, I'm getting so much productivity. But if you look at the whole cycle, it is -- percentages can sometime be misleading. So you have to start to look at applicability in context of how much of that activity in the whole value stream there is an ability to apply these technologies. There are 2 key things which need to be importantly looked at. One is from a dependency. See, you need to still have experience to use LLM and gen AI skills. There's some -- there's whole -- there's a lot of activities to perform. There is maturity of the LLM model, which is actively evolving, a lot of the foundational models being built by these providers. Quality of model training data and then output veracity and changes which are happening. But look at the other aspects, data privacy, copyright, uncertainty and [indiscernible], correct? Production outages, defects, correct? These are areas one will need to start thinking about, accidental use of sensitive data, correct, and then again, today, the way this architecture is, if you roll back and look at, the cloud hyperscalers are the core feeders of Gen AI technology. Look at who is building Gen AI technology. Majority of the work has happened with OpenAI. Obviously, if you look at it, it's all built on Azure. They're using Azure as underlying play. Google, second the Google Cloud with the whole Google AI. And AWS now with CodeWhisperer and AWS Bedrock and also clubbing onto SageMaker and other pieces. So that's what we are following up in this vector. These are the 3 big clouds. So if you really go back to this, it's another layer of consumption on cloud. So you have cloud infrastructure, cloud applications, PaaS capability, you have to bring a lot of data. So -- hence, two things which we believe it's going to happen more and more. That option in enterprise will mirror the cloud journey. So keep the consumer out. Consumer adoption will happen at scale. I just want Google IO, they are releasing a lot of Bard capability inside Android, correct, the next releases of Pixel and other stuff will have so that consumer adoption is higher on at scale, correct? Enterprise adoption because in case of Microsoft, we have to use Azure OpenAI. In the case of Google, you have to use Vertex AI. In the case of Amazon, you have to use Bedrock systems. But actually, if you go back to this, the availability of these systems, like even, let's say, Copilot, now it's getting out controlled introduction releases in different countries. Even in India, for India enterprises, it's still on the waiting list and then as Microsoft opens up, just want to take its time on the time line to get those technologies. Second is you have to bring them into the enterprise training them with your data. So there's a lot of discipline of data engineering, data veracity, data quality, data management. So you need to -- if you want to train the systems with your data, you also need to do a lot of annotation tagging of your data, correct, which data you want to train on. You don't want to just ingest things which you don't want the Gen AI systems to learn or learn and not answer anyone in the enterprise, tenant itself trying to ask the question, correct. And the second aspect is this [indiscernible] discipline is going to become extremely important in cloud as infrastructure became utility. Apps became more and more consumption-oriented. This is actually -- this mythical word called the token, correct? That's what the new pricing model is. You start training language models. We still have not found the right metric of how much it's going to cost and what, it is still an evolving space. So there's going to be a variable cost model that's going to come in for customers. So the FinOps discipline and mirroring the cloud journey. There is a lot of things one needs to factor in because it's moving for a completely variable utility consumption-based model. So what are we doing in this, correct? See, from our perspective, at HCL Tech, we see this as a very interesting perspective. So we see that we have, as a philosophy and as a capability, have been creating, infusing, embedding and integrating AI and Gen AI is a subset of the larger AI discipline. And we've been doing this for the last 2 decades, correct. We've been in from silicon to infrastructure to apps data and processes. And we believe that as we help customers adopt this, that will help them supercharge their business to become a generative enterprise, that -- which is the steep, how do you apply Generative AI in your enterprise capability? So you've got a collection of services, very like a decade plus of work around AI and -- enabling around AI, machine learning, computer vision, robotics, natural language processing, a lot of this work, which have been doing annotation embedded service -- learning services, Trustworthy AI, MLOps, a lot of those pieces. This has been in there. So what's happening and the other huge selection of partnerships from hardware [indiscernible] partnerships all the way up to this, you've got working with the AI ecosystem. But in generative AI, the way we really see this, there are 4 big opportunities. Prompt engineering, I think this is going to be the single biggest demand area because how do you iterate with these systems and tune them, blend them, synthesize them. I mean, just a lot of work, and there's a huge opportunity, which is in there. It needs incremental skill, this thing. It needs a different kind of just like you have Squads which you do in full stack development and teams which you have in operation. You will create a Squad in prompt engineering because you need a variety of different skills to do prompting. It might need a mix of capabilities, including domain and a lot of other pieces. Data engineering, I think, is the biggest opportunity. Creation, capture, cleaning the data, pipelining, data management, data operations and the whole aspect around data, it is going to be huge. See, analytics inside will look most generative because these systems will integrate [indiscernible]. But I think that getting the plumbing right, getting what goes in and what goes out. And along with data engineering, you are all responsible AI, what data, inclusion, fairness, traceability, accountability, trustability of results, correct? There's [indiscernible] change management. There's a lot of other things. If you're using these systems to assist in making decisions, the data, which -- this is exactly the way they say, good behavior, bad behavior, which household the child grew up, which school did this child went to, how well it was groomed and educated, correct? Did they have good or bad influences? So things -- people, just the similar way, generative AI systems will be shaped based on the quality of data which we ingest into it. And the last but very important thing, my favorite quote has always been "God made world in 7 days because he had no installed base." In an enterprise, if you are greenfield, it's completely different. You all live in a world where we have to take all this capability and integrate and orchestrate with them to create intelligent apps, you have to infuse them into your existing landscape. And that's where the opportunity exists from a services perspective, how do you pivot and build those capabilities. What we are doing is that because we have this whole key ecosystem partnerships, especially with the large cloud providers and also with some of the technology creators like IBM, Nvidia, and Intel. We've been working very closing, obviously, with the 3 large hyperscalers, which is AWS, Google AI and Microsoft with Azure OpenAI and the whole collaboration. We've been already engaged with them over a period of time for delivering a lot of engineering services, ISV, OEM and capabilities to them already. And we've been involved in early stage of tech development in partnerships over a period of time. In reverse, we are taking a lot of the technology as a consumer. As HCL Tech, we are consuming a lot of their capabilities, whether it is Copilot, whether it is Duet AI, they're rolling it out within our enterprise, both within our HCL Tech Services landscape and also [indiscernible] software where we are deploying a lot of tools technology. And then we are co-creating a lot of services and capabilities using our gen AI labs, where we can consult, create, infuse and integrate to our end customers. So how do we -- it's again the same thing what we did in the cloud. So this is another workload or another pattern on top of cloud when you apply generative AI to the enterprises. It's another specialized workload. We migrated infrastructure, VMs to cloud, we then containerized, modernized, app modernization, data modernization, be it SAP to cloud. We then started deploying. This is going to be a big data AI pipeline and then you want to deploy all these services, that is Vertex AI. You consume using Google Cloud. Whether it's Azure OpenAI, you need to build on top of Azure. So we clearly see this as an expansion of the Gen AI ecosystem on [indiscernible] hyperscales and it will have a hybrid deployment model. The most interesting thing is going to be -- cloud model is all about bringing data closer to the compute. I think Gen AI will force a little bit of a differential thinking where you have to take compute closer to the data because it's going to be practically impossible. We just bring all the data into one and then there's going to be multiple providers. So it's going to be very interesting how this model is going to evolve. So if you want to -- are you going to put everything around Azure OpenAI, good. But then if you want to start doing something with Google AI, you have to redeploy and recreate and move all the data to the next cloud provider. So there's going to be data duplicity, veracity. But there's a lot of work, so planning, data pipelining, hybrid data, there's a lot of opportunity around that. So that's the way I really look at this. So we have a Gen AI lab, which allows us to quickly ideate, prototype and get this technology from an offering incubation, work with our customers in the edge. Quickly prototype and then mainstream it into scale. It's exactly what we did with cloud native labs 5, 6 years back when cloud native was adopting at scale. So Gen AI lab model is bring the power of what we do in digital business, digital engineering, digital workplace and all our digital process operations and start to infuse this and say, how can we help solve customers' problems -- business problems. We've identified about 150-plus use cases, correct, of what we can apply it. And those -- that's what the Gen AI lab is helping protype. As the customers get access, it's very interesting. We currently -- about when Microsoft released the early preview, they have 60 global customers they -- started to roll it out. A significant chunk of those customers are existing HCL Tech clients. And we've been engaged with some of those -- many of those clients in the early-stage prototyping [indiscernible]. And also, we've got a couple of examples of how [indiscernible] a couple of those applications. But again, as the technology access -- so again, you have to not look at consumer chatGPT and Bard. That's a completely different use case. We have a product use case, I'll talk about that. But in enterprise, they need to get those tenants, correct, can access and get those models in their tenants and then there's a lot of work, but there's a lot of opportunity, which we see in this. Two examples. So one is -- it's a medical devices company where we actually use the chatGPT API and including a mix of different language models to train it as a medical conversational agent. But this is not like a programmable part, but this was like a generative part. You just ingest it. The GPT 3.5-based -- this thing it was initially built on a GPT 3. Previous implementation got upgraded to the newer one. And it can generate based on a context or a conditional prompting. And it helps the agent -- so it's like assisted agent plug-in. So which means that the health care worker uses this to quickly prepare summary for the physician, for efficient treatment, just like a human machine partnership. The second one is a very interesting use case. As we respond to a lot of RFPs and customers, this thing, in various different industries. So there's a unique use case which we've been built like a sales bot, which allows me to do retrieval augmentation, augmented generation, quickly ask the question, tell me by industry, by vertical, by service line, by this thing or by product. Can -- what can I quickly go about a customer, case study, use cases. So it's like rather than searching documents, it can generate like a pre-templatized proposal for someone to really go and start working on. So we built these two as very interesting case studies. We've talked about that productionizing this, but there's a lot about 100-, 150-plus active MVPs, which are in progress, which is currently being done, very interesting use cases, starting from image to code generation to customers. A lot of use cases we're seeing is in customer care, customer service, customer success. Can you create more conversational type engagements around that? So that's where we really see this. So from a takeaway from our services standpoint, I think the six key things which we believe, which gives us a very unique position in this. So one is that we've been involved in creating -- co-creating AI tech stack for last 2 decades. There's nothing new from an AI capability standpoint. Generative AI have some huge potential, correct? And we are an early adopter of this. We are rolling out some of these capabilities, asset class to customer, both the Microsoft stack in our services segment and Google stack on our software products business. We've deployed AIOps in scale, in our business, and that's how we carve some of those IP into our software product business to really monetize into that. We are a large partner for all these hyperscaler programs, so pretty much in the Gen AI stack and then we have spun up the labs, which allows us to quickly iterate similarly on the same journey, which we took our customers to cloud. I'm going to quickly pivot and just give you a perspective on what we're doing in our software product business. A different view. So one is on the services how we create opportunity and how we apply it. And as an ISV -- so our software product is basically built and we offer a lot of our core software products. This is 90-plus products, but our shifting is moving towards what we call the 4 cloud strategy. We have a business cloud, which is the platform of [indiscernible], data cloud, an app dev cloud and an automation cloud. What we're really doing here is working with the 3 large hyperscalers and some specialized niche partners to embed some of this capability, generative AI capability into the cloud. So what we are doing there? So either we are creating some of those models, we are embedding them or plugging them into our products to really fuel our customers' digital plus economy. So 3 things. The first, we are adopting Copilot and Duet AI as a payer programmer to increase our product velocity. So we are seeing how can we cut backlogs, how can we deliver more pipeline, how can we offload some of the mundane tasks of developing applications and building those capability to a Copilot or a Duet AI system and then -- and we have something called AI Workbench, which is called an AutoML platform, which we've already been using for the last 5 years to build all the machine learning models. Now we are plugging that along with this to quickly do model training, MLOps and other pieces. Second is, we just launched last -- 3 weeks back, and this is a large one of the top 5 Indian banks. They are deploying this. They are a big Unica customer who are just rolling this capability out where we are actually, for about a very sizable number of their cards and loans, auto and personal loans, we are using generative AI to do contextualized e-mail generation using our Unica Deliver product and then using Unica Interact to really respond in context. So it's not like a templatized e-mail, which you create in a campaign, send it out, this is going to be -- if the email has to be sent to me, SKK or to Surendra, it will be personalized. If you look up the customer profile database, use the tonality and the language to contextualize a customer. This really scales up their engagement model, correct? So we just released this as part of our marketing cloud. We have a lot of those capabilities coming in. So what we're doing is that we have launched a new engine called [indiscernible] which basically plugs into our multiple clouds, so it's the business cloud, app dev cloud. And basically, it is an abstraction layer, which allows you to interact with multiple language models. You can plug in Vertex AI, Azure OpenAI, public chatGPT, public Bard and any other privately built models, and it can help you to prompt engineering, model chaining, document summarization and cogeneration. But it is infusing generative AI into our products. That's what we've been doing. And interestingly, we own -- the underlying technology for generative AI is vector database, and we actually own a vector database engine in our software product portfolio called Actian Vector. And we are now expanding that to support unstructured and other capabilities. So the end state is that we're going to infuse a lot of this capability into our products. So one of the areas which we're building Gen AI is into automation cloud, where we can move customers from a human-led automation assisted to automation led -- men assisted. So we're creating new subscription licensing models for a lot of our automation products around observabilities, automation, autonomics and all of those pieces. And we continue to start to see this coming in our releases. So that's a nutshell. So in services continue to adopt the technology, use the technology, help the customer deploy and leverage it, and software embed and integrate into our core product line, which allows the customers to leverage the benefits of this. So Surendra, I think the 5 minutes, which I lost in between, I tried to cover it up. But I think, I hope, I tried to address whatever I could and open to any questions. I know you have a lot of questions lined up.

Sure. No, no. So this -- like which is a great presentation, KK. So what I'll do is I have some questions, and I have some questions from clients on e-mail and the webcast. I will try to combine some of them so that like in the interest of time, okay. So let me just start with the first question is, you started by saying that you have been talking to dozens of customers and you have made presentations to a lot of them. So what our customers -- like how are customers thinking about it in terms of readiness, what they want to do. What are the considerations? What are the kind of conversations you're having with your customers at this stage?

It's interesting, most of the customer conversations, which we have, all of them have some sizable cloud provider commit contracts, either with the Microsoft PA or a Google consumption or on AWS. So obviously, they are -- and there's a generative AI capability, which is available on all these three. So they are quickly looking at how can they apply this. So there are 2 things. First is an excitement. There's generally a lot of excitement watching these releases. Then they are starting to say, from excitement, they're doing a lot of ideation to prototypes, so a lot of MVPs, so the Gen AI lab is becoming very interesting. At this moment, our labs is like overwhelmed with MVP requests, like I want to find it quickly, get me something to do, and everyone wants to show something and then when the experimentation starts, you then really say it's okay, then the use cases get to become more visible. Clearly, 3, 4 areas. Correct. One, everyone wants to try this as a conversational -- so customer service, customer care, customer experience, outbound marketing, anywhere there is a lot of one to many conversations around context. About my product, about my offering, about my -- whatever I offer, like if I'm a cards, like you could train the Citi cards basically to -- and train that -- this thing, or any -- any on this, cards business or -- well, so I think they're really looking at -- that's a very interesting use case. Second one is everyone wants to try code generation. I think there will be a lot of people who will try to learn programming now because you can use assist your programming. See, you still need to generate code. But if you don't understand programming, what will you do with the output, you have to still interpret that output, correct? I think -- we still not reach a utopian dream where you could generate a code. It goes into production and becomes an app. It's still -- it's still tied to environment, deployment, data, would still integrate data, a lot of those things. There are some interesting areas around NLP to SQL. I think that's something which we are seeing that. Everyone doesn't need to know SQL. You still need SQL query optimizer and all in the back end. But a lot of people can -- I think that citizen developer, citizen low code has got the huge expansion because as more and more capability comes. That's the second thing. The third thing people are really looking at is this whole area around where you have to look through a lot of documents -- paper documents, knowledge. It's [indiscernible] a lot of our information is in paper, correct? Documents, PDFs, in various formats, ability to ingest that and a lot of us has experienced, remember, we have this stored in this document or we've written something here or your research reports, correct? Like you would have written something, Surendra, you write so much, correct. Suddenly, you have to remember what you wrote 4 years back. But if you ingest a lot of this in a curated way, the right annotation into a generative AI engine, you can ask questions and then you can get. So it's not like just a search and retrieval, but it can also give you synthesis, you want to summarize the document. So there's a lot of those use cases coming in. Then there are very specialized use cases around niche. You've seen a lot of use by UI teams, they're looking at kind of use of the generative UI. So there are some very interesting use cases coming in. Industry sector wise, there's a lot of piloting expansion happening, but very, very important. Customers are now getting those access to those private tenants. There is still a lot of conversations, which I'm having with CSOs and Chief Risk Officers. They're still trying to de-mystify in terms of, okay, this language model training on my data, is the model frozen, will it learn something and go back and update. So there's a lot of those clarity and those conversations are happening. Second thing is, there is suddenly a big uptick in data engineering assets alone because they are realizing that do all these things to need data cleansing. Data foundations have become very important because that's what is going to train the model. So these are the typical conversations we're having with customers. So -- and Europe is a bit different because there's a lot of privacy conversations just triggered up in different, different countries. But that's in a nutshell, the way we are seeing the whole demand pattern.

And from the key consideration perspective, you mentioned the CI -- CSO and the risk guys coming into play. But beyond that, are there other considerations in terms of investments? Or is it still too early to get there? I think it's still in that initial excitement phase I assume?

Lot of prototype and MVP is happening. That's also giving people a lot of areas they need to work on. So the prompt engineering, if you go back, 3 months back, no one talked about prompt engineering. Now there is this new discipline, which is prompt design and prompt engineering, both, correct. The reason is they're realizing if you -- just think the time chatGPT was trained by OpenAI with zillions of prompting and tuning to get it to answer to a certain level, I think the effort you need to put to answer context to our enterprise data, correct? So there is -- that's one. The second thing is whole integration and orchestration because how do you connect? See, chatGPT or Bard is a UI interface. But you will take the UI out of context now and see, okay, when you access them through APIs, how do you push them in? Which language model to choose, correct? Like OpenAI gives you 4, correct? Google gives you 4. Each language model has different costs by token, correct? which one we will use to train the model? Which one we will use to prompt, correct? So these customers are now when they're doing MVP prototyping because even most of the cloud providers have given access to them on a -- like they have a tenant pricing, some of them are early adopter, but they're still not gone into production pricing models, correct? So as they figure the usage out, then they'll have to start thinking about a lot of those pieces. I think it will accelerate a lot of cloud adoption in my view. I think it's going to give a lot of cloud adoption because customers have to move the data to the cloud. So that whole journey to cloud adoption so that you can start consuming generative AI. So that's the reason both Google and Microsoft, if you really see the 2 big push, there is a lot of adoption on that. And obviously, Microsoft has a play on the Microsoft 365, which is around the Teams is Copilot, the Word is Copilot. Similarly, Google on the workspace is doing that Duet AI and other pieces. So there's going to be -- every year AWS is catching up. I think they're coming up and now they're just trying to go hyper speed to just get a lot of their pieces out and some very interesting thing coming out with CodeWhisperer and a lot of the other pieces. So that's going to be the same. I think customers have to then make a choice. Again, go back to the cloud conversation. One cloud on-prem. Now it's going to be multi-cloud. Now the question is how many places will you [ forb ] data, then you will start segmenting saying that, okay, I want to build generative AI for ERP applications, my ERP is running, let's say, on SAP RISE on Azure. I'll surround my data plans around that. A lot of my commerce is running on Google. So maybe I'll not -- so you will start to see those landscape conversations coming in. It's going to be interesting.

Got it. Got it. And the applicability slide that you had was quite interesting. And again, I don't know if I got it right, but just wanted to understand this better. So some of the areas which already had been able to generate significant amount of efficiency using AI in the past, say, BPO, Workforce Planning, infra and data. They are the ones where I think there is a lot more efficiency being generated because of Gen AI. That's what you're expectation seems to be. Is that a correct assessment? Or you think that it's still early and some of those things could change?

Okay. So if you go back to my Slide 1, I always say applicability is a range, correct? And I know it's rather than, say, early, there's lot of potential, correct? So if you really see, there is a lot of potential, correct, in terms of using this technology. As you -- Surendra, again, there are scenarios, correct? Customers who want to -- there are two sides to the story, Surendra. Many times, you are very part of about saying that how will it impact what you do? But I think if you really look at -- you have to start looking at what has the IT services industry done? I think I was having a conversation with a very interesting [indiscernible] like this whole concept. Go back 20 years back, 22 years back, 2-digit to 4-digit conversion, then it became offshoring, then it became Lean Six Sigma, then it became the whole opportunity of [ RIM ], correct, then this whole digital smack came in. It is about pivoting. So the industry has always has to pivot. There are companies which will evolve in pivot and others who will not be able to pivot or will take more time to pivot. So I think the applicability is if you're delivering managed services to customers, so obviously, your ability to deliver better outcomes to the customer isn't there, then you'll figure out what is the right mechanism of sharing those benefits with the customer. You could do more -- your productivity of your developers could increase if you move into a better model. But I think it also has to be contextualized. Let's take an example. Let's say, we are rolling out Copilots and people using Copilot for -- by the way, Copilot is not new. July 22 was GitHub Copilot launch. We have been adopting, at least tens and thousands of our developers have been using it. In our software business, we've been using [indiscernible] programmer for a while in many of the products, correct, where it is contextually, we can use it in a newer programming languages. Now when I use GitHub Copilot, a lot of the code, I can pull out of public data repository. And then I can do it on my own enterprise, correct, which is my own implementation. Now the question which is start to come -- is that customers have to start building those DevOps tools and environment with this capability because it's still that question which you talked about, correct, who owns the copyright, who owns the code if it is generated. There's a lot of things which we'll have to answer. So I think even the customers are trying to pick it up and do it in pockets and segments, so it's going to be fairly -- it's too early for us to just take a universal plate and say, I'm going to -- just going to do this or that, correct? I think in enterprises, the customers will adopt it and ingest in a certain speed. And that's what I think in the slide which I talked about it, you'll mirror pretty much the way cloud adoption, correct, the big [indiscernible] you can look at using this to build something quicker, correct. I also think there's a lot of opportunity in legacy modernization. Trust me, I think there was a statement about -- 6 months back about that 15% of only [indiscernible] yes, 15% of all enterprise workloads are just switched to cloud. There is a huge amount of app mode, which is needed, correct. Could this help accelerate the journey to cloud? The answer is yes, because we are looking at a lot of adoption capability. But if you're building new app, Surendra, you will start to use this capability or you will also use lot of those APIs, which are available for generative AI to start building new applications. But then again, when you start to infuse this into existing applications, you have to open the code, make sure it is put in, tested, integrated, figure out it's working with everything else, then again the big question comes in is, I have enterprise apps, working on secure enterprise data. If I plug in Gen AI, will it work on top of my data. Most of the Gen AI systems do not use an app, they access data directly. You have to feed them data, correct? So then this whole data security, identity management of data is a big, big thing. So there are -- there is a lot of theories that IT industry has to go through in the customer's landscape to make sure it is ready for adoption at scale.

Sure. And again, a couple of questions which I'm kind of combining here. So there is a question on accuracy of these models and how are customers thinking about it? And again, there is another question around, does the importance of testing go up with this kind of model. So I'm just maybe you could cover both of them together?

See, so there's old saying, correct, garbage in, garbage out. This is going to be a bit different. You can import garbage, it can give you intelligent output assuming it looks very clean, but it could be more intelligent garbage. So I would call it because a generative AI, it is all about how you train, what data you put in, correct, hallucination of models, which you've seen in the public world, correct, and I'll give a live example. Before this, I was discussing this with Prateek, correct, our CFO and he's saying, okay, let's go and ask both the engines, correct, the Bard engine and the ChatGPT public engine, hi, tell me what's our stock price yesterday. It's funny enough, you get two different answers, correct. If you ask a question on Monday, one engine thinks yesterday's Sunday, and it knows that the stock market is closed. The other engine just picks up and generates an answer for you, which you think is realistically near. So now the question is, how do you validate it? You then go back to one of those market sites like [indiscernible] Finance, Google Finance, or Moneycontrol and come and say, okay, is this data matching with this data? So I think if you're doing this in a public instance, if you want to test the veracity quality of data, the prompt tuning and this thing, there's going to be a lot of work. So I think it looks very appealing to start with. But when we want to start making decisions or judgment with these systems, the data engineering discipline becomes extremely important. That's one. I think there's a lot of opportunity there. The second question you had was, sorry around the -- Surendra?

Importance of testing.

So test. See -- so one is people say, test case generation is getting automated, it will help you automate. But that is what the test case of the code. Now see the testing of the prompt itself, correct? The more data you create, you have to test it. So there is a lot of opportunity. I think there is some interesting developments happening in the space where the testing of prompt -- automated prompt engineering testing. Some of the work which we are doing in our product [ prompto ] is basically -- is to help that iteration cycle so that you don't need to be a prompt engineer to do prompting, so that we could help you take that, because the volume is going to be huge. Surendra, what we realized is we just picked one use case for a customer in a bank. Our prompts went to thousands and thousands for it to be tuned because the scenarios, the way you ask a question, I ask a question is so different, correct, and I might get two different responses. So the tuning to a baseline model is extremely important. There is a lot of testing, but a different kind of testing. So again, if I do the traditional testing the question that you asked before, correct, if you don't change the way you do things, then you might not be able to address the new opportunity. So I think there is always an evolution here.

Sure. And again, there is a question, and again, you can answer it in the way you want to. Okay. So this question is, while on one hand, you mentioned a lot of opportunities which are coming up because clients are excited and there are so many test cases and so on. But again, some of the examples you gave, right, and you said there are a lot more examples. What are the kind of efficiencies that the coders or the programmers could generate? And again, I understand it's early stage. So this is a question I'm reading out, but you could answer it whichever way you want to.

Yes. See, I think there is a range. See, I gave you a view, correct. It can give you -- see, I don't -- see, there is a lot of research and reports out there, which is saying it's going to give you X or Y. I think in all honesty, GitHub Copilot, which is not the Visual Studio Microsoft Copilot, which is the one -- which they have launched on Azure DevOps, that's a newer one. For certain parts, it is showing some significant value. I think, to be honest, it is too early to make just the prediction. See, I don't want to be saying something which you can't -- I think it's very early stages now, but it can be used, what it is still yet to be proven, easy to take millions of lines of code of an existing system. You have to first train the system. See again, this is not a Copilot use case. You have to first also understand the ingestion of the whole functional specific -- take a very complex trading app, correct, it has so many logic rules, scenarios built in. You have to first understand the application, the way it has been written, their interfaces, the logic which is stored in there. And by the way, not all logic is stored in the code. By the way, a lot of our applications have data logic stored in the databases. So either we still have not reached a point where you have a data layer separation, many of them still -- so I think -- it is -- it looks very easy from outside of generating code because you're building some app on your desktop, doing some nice website design, a few things. Yes, it can generate code, but we are really talking about enterprise systems. It's not working on one, how do you understand the whole system architecture. It is not one product. So to deliver a trading system, you have 20, 25 systems, correct, which goes and processes the whole this thing, how do you build the dependency. So I think in my opinion, if you pick modules, if you pick products, there are areas where it can give you. There are certain areas like, if you want to do random test case generation. Rather than you're thinking about this, you can write some scenarios, you can generate something, but it can help you do certain things, correct? But I think generating code, it does, but I think you still have to validate. So the developer cannot just take the code and deploy it even to a CICD environment without validating and testing. So I think, in my view, practically, if you give -- you have to give it some more time. We've got a lot of examples. I'm not sure I don't want to put out something there. I think let's get it more validated, but I -- that's where we'll be able to give a better output at some point in time.

Sure. And again, maybe we have -- like we're already past the time. So maybe squeezing in last one or two questions. Could you talk about the people and the training aspect of it, how things need to really change?

A few things, correct. I think people have to -- so there are two parts of this. There's a technical skill and there is this whole mindset. I think people have to start to learn how to work with a human machine partnership. So today, you do work with people. You have to start to think how can I have a team or a score, which has got people and Copilots or Duets or assistive pieces. So that's one big thing. So it's about collaborating. Understanding that these systems generate responses where one has to validate and look at that, that's one. Second is that if the skill set needed to data engineering, correct? It is a very, very important skill. So I've been building a skill in training. We're in the process of creating a lot of capabilities around data engineering because -- which is very different than data analytics, correct? Data engineering is the whole data management life cycle, correct, from the ingestion to how we operate the data, correct and how we even retire the data. So it's very important what data you want to get out of the system. I don't want to -- it to be idling around for some data, which doesn't need to [indiscernible] the model, whether it's ML model, whether it is a VRN model or it's a Gen AI model, correct, but that's a skill set. The third thing is this whole area around integration and orchestration will become very, very important skill. Educating people on responsible AI becomes very, very important because as developers, as data engineers, as DevOps engineers, correct, as prompt engineering, as functional people or users, correct, how do you make them aware of responsibility of AI, both in what you put in and how do you take the response back and how do you use it in a more effective way. So that's going to be a very, very important. There's a lot of soft skills along with technical skills, which is needed to be able to use the system. So I think it's going to be -- I think there's a huge opportunity to train just like we did. We train people over a period of time. The whole training engine and services have to now skill people on the newer skill sets which they need.

I can go on and on, but in the interest of time, I'll bring the session to a close, and there are some pending questions, maybe I'll reach out to you via Sanjay. Okay. So we can do that. And thanks a lot for taking time out and doing this session. Obviously, there is a fair bit of confusion among investors. And I think your session was super insightful and helpful in that context. Thanks a lot for your time, and thanks to all the participants today.

Thank you, Surendra, for the opportunity. We look forward to talking to you. Thank you very much. Thanks, everyone. Have a great evening or a day ahead.