EY helps clients create long-term value for all stakeholders. Enabled by data and technology, our services and solutions provide trust through assurance and help clients transform, grow and operate.
At EY, our purpose is building a better working world. The insights and services we provide help to create long-term value for clients, people and society, and to build trust in the capital markets.
Bata India’s CFO explains how AI, analytics and governance are reshaping forecasting, decision‑making and talent to build agile, future‑ready finance teams.
In the second episode of the Finance Transformation podcast, Sardul Seth, Partner and CHS Consulting Leader, EY India and Abhinav Johri, Partner, Technology Consulting, EY India, speak with Amit Aggarwal, Chief Financial Officer, Bata India, about the evolution of finance in an AI-driven world. Amit discusses how integrating advanced analytics and machine learning into finance processes is transforming decision-making, forecasting and customer engagement. He highlights the importance of building AI-ready talent, enabling robust data governance and strategically scaling initiatives to deliver measurable business impact. Tune in to understand how CFOs can lead finance functions that are both future-ready and value-focused.
Key takeaways
Bata uses AI and machine learning to enhance forecasting accuracy, customer insights and operational decision-making.
Transformation efforts emphasize talent development, data management, process automation and scalable AI adoption.
Governance and data security remain critical as organizations leverage AI for finance and business strategy.
Organizations accelerate transformation by developing specialists within their business functions and fostering a culture of experimentation, learning and adoption.
We actively engage our workforce in proof-of-concept projects so they understand the art of the possible in AI adoption.
Amit Aggarwal
Chief Financial Officer, Bata India
Thoughtful leaders propagate a forward-looking data strategy and architecture to enable AI initiatives to be scalable and sustainable.
Abhinav Johri
Partner, Technology Consulting, EY India
Host
Sardul Seth
Partner and CHS Consulting leader, EY India
Abhinav Johri
Partner, Technology Consulting, EY India
For your convenience, full text transcript of this podcast is available below.
Sardul
Hello everyone. Welcome to the EY Finance Transformation podcast, where today we are going to focus on artificial intelligence as a subject. I was reading through the literature and found that 73% of CFOs in the retail space intend to embark on an artificial intelligence journey, but only 18% have acted on this. Therefore, there is a stark difference between reality and execution.
While we were thinking about what to talk to you about today, we thought, why do not we talk to industry stalwarts and have a conversation about what they are doing in their own organizations around this subject.
With that context, I am joined by Amit Aggarwal, Chief Financial Officer, Bata India, a company known to all of us and sells almost 50 million pairs annually in India and has a presence in 1,200+ cities. I am also joined by my partner colleague, Abhinav Johri, who leads our digital strategy and artificial intelligence practice for the Consumer and Health Services market segment at EY India.
Welcome Amit and Abhinav.
Amit and Abhinav
Thanks, Sardul, for hosting us.
Sardul
Amit, let us start with the big picture. Retail is changing. There are a lot of things happening, like omnichannel; it is complex now. With that context, how do you think the role of the CFO is evolving?
Amit
You are right, Sardul. The dynamics are changing from a consumer- preference standpoint and as you rightly mentioned, Bata serves more than 1,200 towns in India. We are a multi-format organization, starting from our own stores, which offer endless aisles and we have a large presence through franchise stores, specifically focusing on tier-three and tier-four distribution.
We have a large distribution and institutional business, plus an e-commerce presence through partner platforms as well as our own website. It has become very complex from a CFO perspective and from a finance function perspective.
From a strategic value-creation perspective, there are three key pivots to which the finance function contributes:
From a traditional P&L perspective, which is the consolidated-level P&L, it has moved to the store level, the channel level and the product level. Those kinds of slices and dice are needed from a business perspective.
Finance drives the entire resource allocation within the organization. Finance is expected to partner in all decisions — what to buy, when to buy and how much to buy, which is core merchandizing and extends to supply chain strategy, eventually translating into consumer sales.
Sardul
That is very interesting. You embarked on value as a lever and theme. Abhinav, we have been advising a lot of organizations around this area, and you have been supporting them not only with transformation but also leveraging data, analytics and artificial intelligence to extract value. What are you seeing in the market? Is it something similar to what Amit is saying, or do you have a different view?
Abhinav
This is quite encouraging, what Amit just explained and how it is being done in Bata. Kudos for the way they have thought through how they run the finance function. From a CFO lens, we have seen that our clients are asking for this.
We are seeing an uptick in how data is being dealt in every organization. Many mature organizations now treat data as a strategic asset. While they continue to deal with their ability to bring data together, they are often a problem due to the integration of a variety of systems that retailers run — progressive technologies now make this a solvable problem.
CFOs like Amit can significantly influence this because they are the consumers of the outcome of the data.
Sardul
How are you keeping the customer as a focus and what are you doing to engage with your customer better? Is there a role that the CFO or finance organization has in that dimension, or is it left only to business, marketing and sales?
Amit
What role does finance play in enhancing consumer experience? Direct influence? No — that is a core job of sales, marketing and the CRM function. However, finance can play a pivotal role in looking at the business KPIs being monitored and how those KPIs can eventually be converted into financial ROI metrics.
Sardul
As we go forward in this conversation, I want to shift the anchor a bit towards GenAI, which I mentioned at the beginning of the podcast. Abhinav, what are you seeing in GenAI — not just in finance, but more broadly in business as a strategy? What are people trying to do?
Abhinav
GenAI is close to reality. We tell our clients: the sooner, the better. There are a few critical areas where we see GenAI contributing and making real results. It builds on previous efforts in data analytics but takes them a step ahead.
For instance, scenario analysis or simulation of risks earlier required teams to wait to run a what-if scenario and then plan mitigation. Today, scenarios can be generated on the run, provided the modeling is done right. That empowers decisions in real-time—or near real-time, which is a great strength for any CFO.
Another area is supplier and market intelligence. Finance is often keen on managing credit and understanding credit risk. POCs are underway through GenAI to know credit risk in real time. Most financial systems that enable forecasting and credit allocation are coming up with GenAI features, bringing these capabilities directly to users without relying on IT dashboards.
Sardul
Listeners would want to know what is happening in the industry. Are businesses implementing this concept? Are there time-tested models? Have you embarked on some of these areas and gained significant value? What are your views?
Amit
Specific nuances to Bata in terms of adoption include the examples you talked about in terms of demand forecasting. Historically, models were based on regression. The problem with those models is that they were static. With the adoption of AI and ML engines running behind it, the accuracy level improves significantly because they factor in the latest trends happening on a day-to-day basis and a continuous learning curve running behind these models, which helps improve forecast accuracies.
Similarly, there are other areas, for example, your entire customer experience, who are your lapsers and how do you target them from a customer retention cohort perspective. There are cases where we use it extensively in our organization.
Sardul
I heard you both talk a lot about what AI can do and what the art of possible is, and I think in your case, Amit, you also shared a few examples of what is happening. One thing that is apparent is that as you move into these models and try out new technology, a new age of risks will unfold. These risks emerge in the models, in the way the data is managed and in the way the organization thinks about its future state. How are you identifying those risks within the context of your business model and how are you taking actions or steps to address them?
Amit
Technology also brings risks associated with it. Starting from the model perspective, you need to constantly check whether the model is developing any bias. You continuously look at the results on both the positive and the negative sides to assess whether the model is giving you a consistent bias or not. Therefore, if there is a bias, you remove that bias so that your outcomes become sharper and better — that is one.
The second risk is on the entire data security and data privacy side, because given the stringent laws — some of which are implemented and others still in the process of implementation — it becomes a huge risk from an organizational perspective. How do you preserve data and ensure its privacy while still reaping benefits from it?
Sardul
Abhinav, do you want to add something around risk, identification and management?
Abhinav
Amit pointed out some of the most pertinent areas, especially the bias in the data and we have been seeing it. There are multiple problems here. You train the model on disparate data sources; they may not have a relationship and may be in a state that has established coherence. Therefore, you end up with a bad model and a bad outcome. So, this remains one big risk.
Therefore, many thoughtful leaders often propagate bringing it all together by means of a data strategy, or putting in place, at a more technical level, a data architecture which is forward-looking. Concepts such as data lakes and delta lakes — we have heard about them a lot.
Second is the third-party AI risk. Amit, you spoke about this for a while — about how buying is not always the best option. We all know that we buy for speed and we build for differentiation. How much you build for differentiation is also part of the strategy.
But imagine that you end up buying a lot of AI tools from multiple vendors; we will end up in a very similar problem we often see in legacy systems, where we do not know who owns the data — whether it is you or the vendor. Who owns the SLAs — you or the vendor? And most importantly, where does the proprietorship of the data and of the model lie — whether it lies with you or with them?
In some cases, it is clear and in some cases it remains ambiguous, but this introduces the risk in most of the AI-ready systems in the market today.
Sardul
There is a requirement for this workforce to undergo a certain degree of training around these new-age topics.What is your organization thinking about that and what are you doing with respect to that in your own department?
Amit
For adoption, we do extensive training of the users in terms of understanding how the new ways of working will look like and how it is going to benefit them in terms of upskilling — from a traditional way of simple invoice processing to a modern way of invoice processing, which is largely driven by a system with minimal or no touch from a human perspective.
People need to be trained in analytical capabilities and have digital savviness. When we are hiring new ones, that is what we are focusing on — people who are good at analytical capability and digital savviness because their adoption and ability to add value and adopt technology will always be much higher.
For the existing workforce, through whatever projects we do — such as a small proof of concept — we actively engage those people responsible for those domains to participate in the project so that they also understand the art of the possible.
Sardul
Yesterday I was talking to somebody and they said they are running maybe ten initiatives around multiple problem statements, but they have a backlog of 150 to 200 problem statements that they want to address. How do you scale?
I agree that all the points mentioned are relevant. But if I am there and have started and I am sitting with lack of bandwidth, availability of talent, or an understanding of ROI, how do I scale? How do I create that culture? How do I create that ecosystem so that most of my organization is thinking or doing something about it on their own? How do you start this whole philosophy?
Amit
Sardul, progressive organizations look at this concern in the following way: if I have 100 business problems to solve and I have already started on the journey, what is recommended is that for each business function, you create a specialist within that sub-function. As an organization, you need to coach them, train them, facilitate the building of their expertise and then help them with the ecosystem of vendors who are capable within those specific domains.
Sardul
Abhinav, AI Academy, does that resonate and make sense?
Abhinav
It all starts from how it was initially thought through. Many years ago, when data became mainstream, thinking for many leaders like Chief Data Officer or Chief Analytics Officer became the role to own this responsibility.
We have continued to see that similar thinking has now been adopted, where a Chief AI Officer is the new persona, we will meet in our day-to-day lives and organizations, who will assume this title. But the expectation remains the same — that he or she sets down the vision through which a POC at scale, or an enterprise-wide AI initiative share the common set of design principles, architectural principles and ensures that the foundational capabilities are not built just for a POC, but for everything in the future.
Sardul
Thank you, everyone. Thank you, Amit and Abhinav, for the great insights. It was knowledgeable for me personally and I am sure you will also enjoy it as you hear this podcast.
While we heard Amit talk about the retail specifics around Bata, this subject is very relevant for all industries. As Amit said, you need to start somewhere and as Abhinav talked about, processes must scale. There are different strategies that people and organizations adopt as they embark on this journey and they continue to evolve in this area.
Thank you for listening to it. Great to have you here. Keep listening and keep looking out for new episodes and new areas.
If you would like to listen to our podcasts on the go:
Drive business growth with technology transformation solutions at EY - integrating tech, operations & architecture to build agile, data-driven organizations
