Saurabh Gupta — President, HFS Research[00:21]
Hey Amar and Pinak, thanks for being here. Let me ask you this. I’ve been hearing so much about how Nissan is moving from incremental modernization to really becoming ready for agentic AI. I want to ask you not about the technology but how is your organizational mindset, architecture, you know, philosophy changing? Because what we are seeing, Amar, at HFS is technology is not the problem. It’s all the paraphernalia around it, you know, our people are not ready, our processes are not ready, our data is not ready. So tell me a little bit about how you’re thinking about it. Now, it’s a great question.
Amardeep Setty — Director – Global IS – Customer Experience, at Nissan Motor Corporation[01:00]
It’s a very good opening question there as well. Basically having an agentic AI or a smart AI on a broken data and the broken process only increases the problem. The way we look at is fundamentally three things. One is the mindset shift I think you touched upon, and also the governance, and finally about the architecture which is needed to support this. In terms of the mindset shift, what I mean by that is, is the data really ready for AI? Is the process ready for the AI to be automated? That’s the first thing that we’re going to generally look at. And the second part, like I said, about the governance, not in the way of slowing down and the compliancy, it’s really about having the right guardrails to connect to the first thing that I talked about, the data ready, can we link it to that? And the process ready, that’s how. And the third one, the architecture. We are using AI itself to modernize the platform. It’s not just about bringing in, not only for the new ones, but even the existing solution, we could not in the past address the technical debt. We are using this to address technical debt and also modernizing platform. These are the fundamentally, these three points will get us ready for the acceleration. That’s how we look at it.
Saurabh Gupta — President, HFS Research[02:24]
That’s fantastic, Amar. And I think the point that you made on AI being used to smash the tech debt is essentially what Publicis Sapient and HFS have been talking about now for almost a year, on AI on one hand is exposing the tech debt because you have to solve for it. But on the other end, there’s a flywheel effect where I can use it to modernize some of the tech debt as well. But Pinak, let me ask you, you know, Amar made some fantastic points on mindset, on data organization, on IT architecture. How are you at Publicis Sapient trying to help clients on that? Because these are not your traditional IT services.
Pinak Kiran Vedalankar — Group VP Technology & International Head of Engineering at Publicis Sapient[03:05]
I think the first part is, like what we believe is, I think there’s a lot of focus on tech debt in general. And we have done the paper together on that and the likes, right, basically. But I feel the next wave of the debt that we all need to deal with is a more context debt, right? Where most of the enterprise, if you look at them, they don’t operate on rules and policies. They operate on exceptions to those rules and policies, right? So, and that’s why if you look at the contact committee, steering committees and service introduction, everything is based on the exceptions part of it and the stuff, right? And most of those decisions are made in like Excels and PowerPoints and Word documents, right? Whereas it never kind of gets together as an enterprise decision, which is kind of locked somewhere more centrally, right? And that’s where what we fundamentally believe is the heart of eradicating this tech debt or going into this modernization wave is having this enterprise context graph centrally created for the organization. And this is where what we focus on is get the entire organization, do an AI driven ontology, define different entities, define the relationship between the entities, have the entire embeddings in the vector store for those entities available. This is where you start logging those decisions, logging all the exceptions to the rules, right? So if I give a very simple analogy, if you, let’s say, walk into a restaurant and there is a kitchen where a chef is making a food, there are things like utensils, the recipe book, your oven, microwave, all of that are effectively your system of records, your SORs, right? But it’s just static, it’s there kind of stuff, right? But now, chef never makes the food in the same way, in a way, right? I mean, let’s say there’s a shortage of a specific ingredient. You will make a decision. Oh, I need to kind of now do this differently because I have the shortage. The weather outside is different. Oh, you know what? I’m going to make this thing. The customer has a specific requirement. Oh, no, let me kind of go and do this kind of stuff, right? So that is effectively, I write all of the chef’s notebook.
Saurabh Gupta — President, HFS Research[05:11]
Yeah.
Pinak Kiran Vedalankar — Group VP Technology & International Head of Engineering at Publicis Sapient[05:12]
Right? And how do I now codify the chef’s notebook into this enterprise context graph to kind of add real value for the enterprise? That’s our primary focus.
Saurabh Gupta — President, HFS Research[05:19]
And that’s the reason why I can never cook like my mom. Exactly. Because we all follow recipes, basically. Amar, tell me about this partnership that you have with Publicis Sapient. And what are you guys doing together? How are they helping you? And how’s it going?
Amardeep Setty — Director – Global IS – Customer Experience, at Nissan Motor Corporation[05:38]
We have been working with Publicis Sapient nearly 10-12 years. And most recently with the AI and the agentic AI coming in, the whole model and the industry, it’s not about just having Sapient’s developer resource and the architect to be working with us. How do we bring this whole new way of thinking? How do we address this? So fundamentally, we’re looking at with Sapient three areas. Again, one is in the how we develop, how we bring a consistent platform, and how we operate. In my specific remit, in Nissan, we’ve got various AI-based initiatives we are looking at. However, with Publicis Sapient, it’s very specific, very consciously into marketing and sales and customer experience area. That’s what we do. In the context of development, we use Publicis Sapient’s Slingshot to optimize our whole way of developing, designing, business analysis capture and automated testing and all of that. That’s one area. And the second one, like I said, we have many globally, we’ve got many agentic initiatives and the even siloed initiatives. We’re going with a platform mindset globally. With that, we are embedding both the platform centrally and that is Sapient’s agent platform. Using that, how do we provide a platform globally for all the markets so that they can concentrate on the customer-facing solutions rather than trying to build this agent multiple times. And finally, like I said, operate. How do we operate? We’re again using what’s called SustainAI by Sapient. In the past, it was all about manual way of ticketing and dealing with the tickets. How do we fix it and operate? Now with the real-time observability data coming from the platform and historical data of how these issues used to be fixed. All of that with SustainAI, we are bringing cost efficiency, which is given. It’s not just about cost efficiency, actually improving the response time and the SLAs and all that. So these are the three fundamentals we are working with Sapient and this in the customer experience area. I would say, yeah, it’s going quite well. Love it. You summarized it much better than me.
Saurabh Gupta — President, HFS Research[08:08]
Yeah, no, no. I think that is fantastic. And I think we fundamentally believe, Amar, at HFS, that IT services for tomorrow will look fundamentally different than what it used to be for the last 30 years or so. And I love your Slingshot. You know, if nothing, that’s a great brand name. True, right? It’s a very good name. It is a slingshot. But look, let me turn our attention to the last three hours that we’d be logging on this round table with 20-odd enterprise leaders on tech debt, etc. You know, we all know that, you know, if you look at the global 2000 enterprises, they bought a $1.5 trillion tech debt, and it’s only rising. And we’ve been trying to modernize ever since I joined this thing, you know, for the last 30 years. And it’s like the Windows-Microsoft loading, right? The thing keeps rotating, but it never actually ends up there. Let me ask you both and maybe Pinak, I’ll start with you. Do you see light at the end of the tunnel with AI coming in on this, resolving this tech debt or are we going to have the same conversation when we’re 70 years old?
Pinak Kiran Vedalankar — Group VP Technology & International Head of Engineering at Publicis Sapient[09:20]
I think AI will fundamentally help us transform. Like fundamentally, like deep down, my belief is that it will help. I think from, from at least our side of things and specifically you mentioned Slingshot, right? Basically, and then tech debt, if I can correlate those two, right? The three focus we have, right? I mean, other than the enterprise context that I’ve been in the heart of it, right? I think the three of the focus we have, which helps in eradicating the tech debt completely and which is very, very imminent, right? The first one is how can we help solve higher order of problems for the enterprise, right? I think it is very easy to start solving like a coding problem or a testing problem or something kind of stuff, right? I mean, there are like thousand tools available in the market which can solve that, right? But our focus is like, how do we help in portfolio analysis? Discovery, service introduction, non-functional requirement like performance engineering, operational readiness, like those dimensions is like our key focus, solving those higher order of the problem. The second big focus we have as part of the platform which will help substantially in eradicating the tech debt is improvement in accuracy. A lot of people today I’ve seen go after the efficiency and productivity side of things, like 30% efficient, 40% efficient, 50% efficient, etc. That’s great, but if it is not accurate, it cannot be efficient as such, basically. In our experience, even if you improve the accuracy by 1%, your productivity jumps by 10% kind of stuff, right? So our focus is how do we improve the accuracy and get it first right time from an AI model, stop it hallucinating, etc, etc, kind of stuff, right? Basically, having the context to kind of do that. The second one, and the third one is what we believe is like a one-pass debt elimination, which is rather than trying to eliminate the tech debt, get to the basic and then modernize, we are like, we don’t want to replicate your old mainframes into the new world, right? So rather than a batch process in a mainframe, I don’t want a batch process in Java, right? What I want is an event-based Kafka streaming architecture, right? If something is non-modular, non-reusable over here, I need a composable, modular, reusable architecture, right? So how do I kind of focus on one-pass debt elimination, goes from the current, basically like a genuine target architecture. Those are the three big focus point and that’s why I feel I think AI will substantially help us in this entire dimension.
Saurabh Gupta — President, HFS Research[11:33]
Amar, are you as optimistic? I think I like the way Pinak summarized it actually in the sense
Amardeep Setty — Director – Global IS – Customer Experience, at Nissan Motor Corporation[11:39]
there is no point in, like, it sort of links back to my first point as well, there is no point in just automating the legacy process and the legacy data and trying to solve that. We need to fundamentally look at that differently. And in terms of the optimism wise, I think there is new hope for sure, where we could not make a business case, that was really, really hard. And then always a question of what’s the business value. Now with the AI coming into play, it becomes really something tangible. We can actually make a business case relatively easily. That’s one thing I would look at. But reality is, in the process, will we introduce more tech debt? I bet we will introduce a lot more tech debt. In that sense, I think when we become 70, we’ll have new problems to solve. I think so.
Saurabh Gupta — President, HFS Research[12:26]
This has been a fascinating conversation. So thank you both. I think it’s an ongoing battle that we’re all fighting, whether you’re a service provider, whether you’re a client, even whether we’re an analyst or an advisor. I think who knows what the future holds for us. But I think all three of us can agree that there is light at the end of the tunnel. So on that note, thank you so much for spending some time with us.
Pinak Kiran Vedalankar — Group VP Technology & International Head of Engineering at Publicis Sapient[12:48]
Excellent. Thank you. Thanks, Q.
