AI Transformation in Practice_ Insights from India’s Consulting Leaders

20 Feb 2026 10:00h - 11:00h

Session at a glance

Summary

This discussion featured a panel with consulting leaders Romal Shetty from Deloitte and Sanjeev Krishan from PwC, moderated by Vedica Kant, focusing on how AI is transforming the consulting industry and enterprise operations. Both speakers emphasized that AI represents one of the most disruptive technological shifts in a generation, requiring organizations to reimagine their business models rather than simply optimize existing processes.

Shetty explained how Deloitte is inverting its traditional pyramid consulting model from “1 client to 10 people” to “10 clients to 1 person” by leveraging AI and digital tools, enabling them to serve previously untapped markets like India’s 75 million MSMEs. He provided concrete examples of AI implementation, including tools that save 60,000 hours in audit confirmations and platforms that allow SMEs to create digital marketing campaigns in five minutes using simple language prompts. Krishan highlighted PwC’s billion-dollar AI investment and the development of internal tools like “Chat PwC” and “Navigate Tax Hub,” emphasizing that AI should be viewed as a utility that enhances human capabilities.

Both leaders acknowledged significant challenges in enterprise AI adoption, with Krishan noting that only 12% of corporations report achieving both top-line and bottom-line benefits from AI investments. Key obstacles include data governance and security concerns, potential “bill shock” from token-based pricing models, and the perennial challenge of change management. The speakers addressed concerns about AI commoditizing consulting services by emphasizing the need to move up the value chain, focusing on creative problem-solving, judgment-based work, and human-centered skills like empathy and critical thinking.

Regarding workforce transformation, both leaders acknowledged that the consulting pyramid will reshape, with some middle management roles shrinking while new skill requirements emerge. They stressed the importance of reskilling employees and democratizing innovation, allowing practitioners rather than just technical experts to build AI solutions. The discussion concluded with optimism about AI’s potential to enable entrepreneurship at scale and create new market opportunities, particularly in government technology and emerging economies, while maintaining that human oversight and creativity remain irreplaceable elements of successful consulting services.

Keypoints

Major Discussion Points:

– AI’s transformative impact on consulting business models: Both speakers discussed how AI is fundamentally reshaping the traditional consulting pyramid model, with Romal Shetty explaining how they’re moving from a “1 client to 10 people” model to potentially “10 clients to 1 person” through AI automation, particularly targeting previously underserved MSME markets.

– Internal AI adoption and workforce transformation: The discussion covered how both firms are implementing AI tools internally (like PwC’s “Chat PwC” and Deloitte’s confirmation balance tools), while grappling with questions about how the workforce pyramid will reshape, what skills will be needed, and how to communicate these changes to employees.

– Enterprise AI implementation challenges: Both leaders acknowledged that while AI shows promise, enterprise adoption faces significant hurdles including data governance and security concerns, change management issues, pricing uncertainties around token systems, and the fact that only 12% of corporations report seeing both top-line and bottom-line benefits from AI investments.

– Pricing pressures and value proposition evolution: The speakers addressed concerns about commoditization of consulting services due to AI, discussing how they’re moving up the value chain, shifting toward value-based billing rather than time-and-materials models, and partnering with AI companies rather than competing directly with them.

– Future of education and skills development: The conversation touched on how educational systems need to evolve to prepare students for an AI-driven world, emphasizing critical thinking, judgment capabilities, and the ability to work alongside machines rather than traditional rote learning.

Overall Purpose:

The discussion aimed to provide insights into how major consulting firms (Deloitte and PwC) are adapting to and leveraging AI technology, both internally and for client services, while addressing concerns about industry disruption and the future of consulting work.

Overall Tone:

The tone was pragmatically optimistic and refreshingly candid. Both speakers were honest about challenges and uncertainties while maintaining confidence in their firms’ ability to adapt. The conversation was professional yet accessible, with speakers avoiding overly technical jargon and providing concrete examples. The tone remained consistently balanced throughout – neither dismissive of AI’s disruptive potential nor overly alarmist about threats to the consulting industry. The audience questions added energy and specificity to the discussion, maintaining engagement throughout the session.

Speakers

Speakers from the provided list:

– Vedica Kant – Moderator/Host of the panel discussion

– Romal Shetty – Representative from Deloitte (consulting firm leader)

– Sanjeev Krishan – Representative from PwC (consulting firm leader)

– Audience member 1 – Founder of Corral Inc

– Audience member 2 – Abhinav Saxena, Consultant at Capacity Building Commission, Government of India

– Audience member 3 – Student

– Audience member 4 – Geeta, from GCC (Global Capability Center) background

– Audience member 5 – Sudhakar Gandhey, Former Senior Director at American Express Bank, built Access Cadets Technologies (a $100 million technology company)

– Audience member 6 – Role/title not mentioned

– Audience member 7 – Piyush from Digivancy

Additional speakers:

None – all speakers identified in the transcript were included in the provided speakers names list.

Full session report

This comprehensive discussion featured consulting leaders Romal Shetty from Deloitte and Sanjeev Krishan from PwC, moderated by Vedica Kant, exploring how artificial intelligence is fundamentally transforming the consulting industry and enterprise operations. The conversation provided candid insights into both the transformative potential and practical challenges of AI implementation across professional services and broader business contexts.

AI as a Fundamental Business Model Disruptor

Both speakers positioned AI as one of the most significant disruptive forces in a generation, requiring organisations to reimagine rather than merely optimise their operations. Shetty articulated perhaps the most striking transformation concept, describing how Deloitte is inverting its traditional consulting pyramid model from “1 client to 10 people” to potentially “10 clients to 1 person” through AI automation. This inversion enables consulting firms to access previously unserviceable markets, particularly India’s 75 million MSMEs (Micro, Small, and Medium Enterprises), which large consulting firms have historically been unable to serve due to economic constraints.

The business model transformation extends beyond simple efficiency gains. Shetty emphasised how AI is democratising innovation within Deloitte, where practitioners rather than technical experts are building tools that save thousands of hours. One audit tool developed by a practitioner saves 60,000 confirmations quarterly for large clients by automating balance confirmations from banks, debtors, customers, and vendors. For MSMEs, Deloitte has developed a digital marketing platform that creates campaigns in 5 minutes using simple prompts in multiple languages.

In consulting work, Shetty provided a detailed example of an automobile manufacturer building a plant in Karnataka to manufacture a car every 2 minutes 32 seconds. Digital simulation tools helped identify potential robot clashes and material flow challenges before physical implementation. This concept extends to other sectors—ICU placement in hospitals and Jaguar jet aircraft simulators—representing new types of consulting work enabled by AI.

Krishan offered a complementary perspective, viewing AI as a utility that organisations must embrace, with the differentiator being what they create with it rather than mere access to the technology. PwC’s commitment of almost a billion dollars to AI platforms and upskilling demonstrates the scale of investment required for meaningful transformation. The firm has implemented “Chat PwC” across all PwC personnel, enabling employees to work with AI for efficiency gains whilst inspiring client-focused solutions.

Internal AI Implementation and Workforce Transformation

The discussion revealed sophisticated approaches to internal AI adoption that go beyond simple automation. Both firms emphasise human-in-the-loop implementations, recognising that whilst AI can handle 80% of certain tasks, human judgment, creativity, and empathy remain crucial for the remaining 20%. This approach addresses both quality concerns and the need to maintain human oversight in professional services.

The workforce implications are complex and nuanced. Shetty acknowledged that the traditional consulting pyramid will reshape, with middle management potentially shrinking in some areas whilst new opportunities emerge in previously unserviced markets. However, rather than simple job displacement, the transformation requires different skills: critical thinking, judgment capabilities, empathy, and the ability to orchestrate multiple AI systems and data sources.

Krishan admitted uncertainty about the exact future shape of the consulting pyramid but emphasised that the type of people hired will be fundamentally different. Associates may perform work traditionally done by managers, whilst everyone spends more time validating assumptions, simulating scenarios, and engaging in high-value strategic work rather than data cleaning and routine analysis.

Enterprise AI Adoption Challenges and Market Reality

Despite the transformative potential, both speakers provided sobering assessments of current enterprise AI adoption. Krishan cited PwC’s global CEO survey launched in January, showing that only 12% of corporations report achieving both “vanity” (top-line) and “sanity” (bottom-line) benefits from AI investments, despite significant spending.

The speakers identified several critical barriers preventing AI pilots from scaling to production. Change management emerges as the primary obstacle, with Krishan observing that “humans oppose change, whatever that change may be, even though that change may be invented by them.” This human resistance to change, rather than technical limitations, explains why many AI initiatives fail to move beyond sandbox environments.

Data governance and security concerns present another significant barrier. Shetty shared a compelling anecdote about an aerospace company discovering their proprietary designs appearing in ChatGPT, despite never using the platform themselves. Investigation revealed that vendors were uploading client designs to ChatGPT during RFP processes, highlighting the complex challenge of maintaining IP protection in interconnected business ecosystems.

Financial sustainability also poses challenges. Shetty warned of potential “dramatic bill shock” when AI pricing reaches market rates, creating uncertainty about which solutions to implement and when, leading many organisations to delay production deployments.

Addressing Disruption and Competitive Threats

The conversation directly confronted widespread concerns about AI commoditising consulting services. Both speakers acknowledged that commoditised work will disappear but argued for proactive self-disruption rather than defensive strategies. Shetty emphasised the importance of cannibalising existing services before competitors do, whilst maintaining realistic expectations about transformation timelines.

Krishan positioned this challenge within a broader context, noting that the key lies in focusing on value creation and protection for clients, addressing complex challenges that require contextual understanding of generational issues, succession planning, technology integration, and geopolitical disruptions.

Rather than competing directly with AI technology firms, both speakers advocate for strategic partnerships. Krishan highlighted that PwC was the first to partner with Harvey (OpenAI-funded) for tax and legal work and their recent collaboration with Anthropic. These partnerships leverage the consulting firms’ extensive client relationships and domain expertise whilst accessing cutting-edge AI capabilities.

The speakers also argued that complex enterprise transformations require multiple specialisms working together, providing consulting firms with defensive advantages against pure-play technology companies entering the consulting space.

Educational System Reform and Future Skills

The discussion highlighted fundamental misalignments between current educational systems and future workforce needs. Krishan noted that 95% of engineering curricula remain unchanged from 25 years ago, despite massive technological advances.

Future skills requirements emphasise working with technology rather than coding, with increasing importance placed on psychology, sociology, and human behaviour understanding. The shift from rote learning to conceptual understanding becomes crucial, as AI can handle information processing whilst humans focus on creative application and judgment.

Shetty made a philosophical point about AI’s limitations, noting that AI is based on past inferences: “AI couldn’t have built Aadhar originally, but could suggest it now.” This highlights the continued need for human creativity and original thinking.

Government Technology and Societal Impact

The GovTech sector emerged as a significant opportunity area for AI-enabled consulting services. Shetty provided examples of government applications, including helping chief ministers assess infrastructure costs through geospatial analysis and AI, and improving MSME access to credit by leveraging data analytics to reduce borrowing costs from 24% to 8-9% through better risk assessment.

Krishan positioned AI as a potential enabler of entrepreneurship at scale, drawing parallels to how UPI enabled financial entrepreneurship in India. An audience member who was a government consultant shared experiences of state-level AI tool calibration that proved chaotic but ultimately successful, highlighting the importance of change management in public sector AI deployments.

Market Dynamics and Financial Sustainability

The discussion addressed concerns about the financial sustainability of current AI investment levels. An audience member, identified as the founder of Corral Inc, raised questions about potential market re-rating, noting that major technology companies are raising unprecedented amounts of debt to fund AI development. Both speakers acknowledged that market corrections are inevitable, with some companies succeeding whilst others fail—a natural pattern in disruptive technology cycles.

Shetty emphasised the importance of avoiding both extreme optimism and pessimism about AI’s timeline and impact. Whilst transformation is inevitable, the pace and specific outcomes remain uncertain.

Future Pathways and Strategic Recommendations

Both speakers advocated for India developing its own AI pathways rather than simply replicating US business models. An audience member noted that India is third in the AI race after the USA and China, suggesting opportunities for unique positioning. India’s strengths in scaling solutions and serving diverse markets could create unique opportunities in AI applications.

For consulting firms, the path forward involves embracing AI as an enabler whilst focusing on uniquely human capabilities. This includes creative problem-solving, contextual understanding, empathy, and the ability to orchestrate complex solutions across multiple domains.

The conversation concluded with optimism about AI’s potential to create abundance and societal impact, whilst acknowledging the significant challenges in realising this potential. Success will depend on thoughtful implementation, effective change management, and maintaining focus on human-centred value creation rather than technology for its own sake.

The speakers’ candid acknowledgment of uncertainties and challenges, combined with their strategic optimism, offers a realistic framework for understanding AI’s impact on professional services and navigating the complex interplay between technological capabilities, human factors, and organisational change management.

Session transcript

Vedica Kant

I think we are capped by time to a slightly shorter session today, but we’ll aim to get the most out of it, and I’ll open up to questions as well. I’d like to start off with a couple of common questions to both of you, just to get both your perspectives. I think one is to start with this question of, you know, what does AI mean for you internally? Would love to hear from you each. When it comes to using AI within Deloitte, within PwC, what are you seeing in terms of workflows, in terms of use cases, where you’ve really seen AI already move the needle for your organizations? I think it would be great to hear a couple of tangible examples.

I’ll start with you.

Romal Shetty

Thank you, Vedika, and good afternoon, everyone. It’s lovely to be here on this panel. For us, AI is, I mean, it is, and it is true that this is one of the most disruptive things that have happened, and it happens in a generation. Or more than a generation, something like this comes up. and what it means for us is to really, for us and for our clients, is to reimagine everything possible because this is the one part. AI can do a lot of optimization, but reimagination is an important part. And I’ll give you an example of, you know, because most people have predicted the demise of all of our firms, so it’s always good to hear when people talk about our early demise.

But how we’ve thought through this is part of AI is to relook at our business model. Our business model, largely in consulting, largely in consulting is a pyramid model, right? It’s one client, 10 people, that sort of the model. But if you really look at now, and we large firms, largely, we don’t service today probably the MSME as a segment. You know, we generally tend to do the top Indian corporates, the large multinational companies. But with the ability to have today generative AI and agent tech, and build it and combine it with digital, you can actually invert the business model of, you know, 1 is to 10 to 10 clients to 1 person, where 80 % is done by a machine, 20 % is done by a human being.

So really something for us, which, so we are going to access a market which we could have never done, right? So that is one part of it. The second part of it is to figure out everything that we do, can we do some things faster? To give you an example for in our audit business, in our audit business, we have something called confirmation of balances. That really means that, you know, you need confirmation from your bankers, from your debtors, from your customers, vendors, you know, so that your financial statements are properly stated. For some large clients, this could be like 50 ,000, 60 ,000 confirmations on a quarterly basis. So now, you know, we have actually built a tool, and built a tool not by an expert in tech, but a practitioner where we have democratized innovation.

where that individual now can save 60 ,000 hours for us so that we can spend a little bit more time on judgment -related matters. That is the second part. Third is just to bring in tax. I’m giving you different examples. In tax, to basically say that, can I give tax opinions now much faster by using Gen AI? Fourth, in terms of consulting, to say that, I’ll give you a classic thing. You have a large automobile manufacturer in the world. Who is building a plant in Karnataka where they will manufacture a car every 2 minutes 32 seconds. Now, what’s interesting is, when you digitally simulate this, you’re able to tell the automaker that your robots will actually have clashes, your kinetics will be a challenge, and your material flow will be a challenge, and therefore you cannot manufacture in 2 minutes 32 seconds.

Therefore, redesign your factory in this way. What’s interesting is that conceptually, this can be now taken to, hospitals, where you can say that in an ICU, where do you place the ICU in the best possible way so that there is absolute easy movement of patient flow. So we’re building simulators for the Jaguar jet aircraft. Now, if you said consulting companies would be building Jaguar jet flight simulators, that wouldn’t have happened and in 40 days. So our business models, the kind of work that we actually do, reimagine things for clients and of course within our bringing in our productivity. So all of that has actually helped from an AI perspective. And of course, you’ve got to be careful that there has to be a human -led or human in the loop because you can end up with some serious challenges as well.

Vedica Kant

Touch on some of those challenges and the implications of the use of AI. Sanjeev, good luck for you to chime in.

Sanjeev Krishan

Yeah, so once again, good afternoon and thank you for having me. See, I mean, you know, I look at AI as more as a utility, you know, and it’s something which most of us will embrace. The question is, what can we make out? of it. And that would be the differentiator from a value perspective because that’s what, because we speak about how consulting firms are going to deal with it. And that’s why I mean, if I were to go back in time in 2023, actually, I think we were amongst the first ones to actually commit almost a billion dollars to AI at that point in time, and that was a platform discussion that we had with one of the hyperscalers.

We also focused on, we also committed a significant amount of money for upscaling our people at that point in time. And I think that’s been a key driver for us that, you know, it’s there, it’s here to stay. What do we make out of it? And how do we make sure that we are working with it as opposed to necessarily trying to say that, okay, you know, we are working against it. That’s the first part. So the first part is adoption. And within the adoption journey, let me just say that, you know, now today, for instance, I would say all PwC personnel across the board would have access to what we call chat PwC. You know, which is where we work with AI in some ways to create efficiency, et cetera, et cetera.

and I can say that the human part is something that we at times miss because who’s using it? My people are using it, our people are using it and when they use this, they are the ones who actually came up with multiple things that they could do with it and that inspiration caused us to come up with, I mean, you know, just as an example that Romer gave, I would like to give a tax example, where they said that the manifestation of what they have seen with Chad PwC and others is to come up with how they can solve client problems, the ones which are the most sticky and that got us to actually come up with Navigate Tax Hub which is an AI -driven tax tool that we came up with which we launched about six or seven months back.

Now, let me tell you that it is the people who actually said that, okay, we want to work with it for 12 to 15 months before you actually take it to market and I think that’s how making sure that AI is one being leveraged, you work with AI, you get your people to embrace it. then I think automatically the outcomes for your clients and others will come through. And we can talk about multiple use cases. But I want to really say that it is about us embracing AI, working with it. The value that will come of it will be immense.

Vedica Kant

I want to touch just a follow -up question. You talked about the pyramid within consulting and the impact that AI has on productivity. I mean, as a consultant myself, I know that these conversations about how the pyramid is going to get restructured potentially are top of mind for all consulting leaders. How are you thinking about that? Do you see the pyramid becoming more distinctly shaped, a different shape, so to speak, where you have senior leaders and then fewer middle management, but then more junior people who are able to work with AI? And so that’s one question. How does the shape of the firms change? And the second question is how are you also communicating it to your own people?

I know the big four in India have a very, very large talent pool here. How are those conversations going?

Romal Shetty

Yeah, so we’re re -looking at every aspect of what we do and what that means at an entry level, middle level, and at the top level. And you’re right. So in some parts of it, it’s a clear indicator that the middle actually shrinks a little bit. In some part of it, it’s the juniors that actually get impacted. But the way, Vedika, I was looking at it is one part is this is the business of today. When I spoke about the MSME business, to give you a sense, there are 75 million MSMEs. I don’t service anybody or don’t service much just from a dramatic impact. If I service even one million MSMEs with the inverted business model, I need a lot more people and slightly different skills of human working with the business model.

So I’m working with the machine, having some critical thinking. judgment capabilities and also having a little bit of empathy as well. So I think that’s how we are re -looking at our workforce to bring in some of those skills which were not something that earlier that we looked at. Now, if you look at coding, coding can be 80 % done faster. But then I, when I look at a lot of what is being done in AI is all based on past inferences. Can, could AI have built an Aadhar? The answer is no. Today, can Aadhar suggest, can AI suggest an Aadhar? Right? It can. But it couldn’t have built something new. So can we be creative?

And I’ll give you another example of digital marketing. We’ve built something where, again I’m just taking MSMEs just as a common theme. They never could brand or market their products. We’ve created a platform today where in five minutes, you can actually have campaigns across Insta, across LinkedIn, across various social media channels, digital campaigns, by simple prompts. You don’t need to understand Java or anything else. You just need to know English or Hindi or any other language, Bhashani, any language that Bhashani will support. that’s all and you can actually have campaigns running so it’s about how you relook at your market size and scale how do you skill your people in today and you do reshape and it’s not one size fits all that this is exactly the pyramid model this is exactly the cylinder model it does vary depending on sometimes sector sometimes competency I’m

Sanjeev Krishan

I think since you asked the question about pyramid I mean honestly I don’t know the answer to the pyramid question all I would say is that I do believe however that the kind of people that we would hire would be very different our expertise is the client base that we have which is far beyond that any other firm could expect to have and the domain that we have and I don’t think those things go away so and also you know what is it that as I said what is it that whoever is there will do with the AI right I mean whether it is somebody on the manager level associate level whatever certainly I would expect the work of a manager today to be done by an associate or a senior associate and so on and so forth.

And hopefully they’ll be skilled enough to be able to do so. But I think the critical point for me is that you end up spending a lot more time not cleaning data, but making sure that you are validating multiple assumptions. And then you are actually simulating those to come up with, you know, potential hypothesis for your client and then actually getting into the execution of it once you have made a suggestion to them. So you are far more engaged. And that, I believe, will help us retain value, right? Because you know, I see a lot of work that we do currently could be data cleaning work. Maybe that will go away. But I do believe a lot of highly value -accredited work will come in.

And we will certainly need to have a different workforce.

Vedica Kant

A kind of different angle and a question to you. You know, you talked about how AI has impacted some of your work internally. We’ve when it comes to clients, we’ve recently seen a lot of studies which say, yes, AI is is great, but when it comes to an enterprise setting, it’s perhaps not giving the same kind of ROI that people expected. And enterprises are complex. Workflows are complex. I would love to hear from you, what are some of the challenges that you’re seeing when it comes to deploying AI in enterprises? And do you see that as just teething troubles? Do you see it as something that is just part of how enterprises work, so it’ll always be complicated?

We just love your perspective there.

Sanjeev Krishan

I think the problem is that humans oppose change, whatever that change may be, even though that change may be invented by them. So I think the problem is not with intelligence. It is about the change management and the integration pieces of it. And I do believe in every organization, whether a consulting organization or otherwise, there will be challenges when people are asked to adopt a particular use case, assuming that it has had success. and I think we will not be any different. I’m sure for us also, it will be a challenge. For our clients also, that will be the challenge. That is why you see a lot of people getting very happy with some pilots or doing some sandbox arrangements, etc.

But when you want them to scale, it becomes different because adoption and integration of that, the change management piece is the one that I think we haven’t even started testing it, to be honest. And possibly that is the reason I’ll be short here that when we actually launched our global CEO survey, just in January last month, it just said that only 12%, only 12 % corporations, in spite of having spent some money, or I would say significant amount of money, are saying that they have got both vanity, which is top line, and sanity, which is bottom line, through use of AI. Only 12%. So I think we have a way to go.

Romal Shetty

I agree with Sanjeev. I think just a couple of other points. Why are pilots not getting into sort of really, really production -grade? One is the governance over my data and security. I’ll give you an example. An aerospace company said suddenly they saw that their designs coming in chat GPT. Now, they say that they have never used chat GPT at all. So where are the designs coming into chat GPT? What they realized is when they were doing RFPs for their vendors, right, and they would give some designs, the vendors were uploading it in chat GPT to figure out a solution. So how are you actually managing your data and IP? Because if everybody uses AI, what is your IP?

So that’s the first one. The second one is everybody’s understanding in terms of tokens. Now, if you take the telecom parlance, you know, when 2G, 3G, 4G, 5G happened, you saw tremendous amount of data being downloaded with 5G, you know, because it was like a free -for -all and the price has gone down. Today, the way the token system is that you love it, and so you keep using as much as possible. but they are all subsidized today. The day this happens where they bring it to some reasonable price because everybody has to make money someday, there will be a bill shock, dramatic bill shock. So I think if you look at some of these aspects and third is, you know, again, new technologies coming again and again.

People don’t know, should I wait? You know, something else is coming. So should I then sort of implement that? So there is a bit of confusion and how does all this orchestration work? Five different things. So I think an adoption, I mean adoption and change management, whether with technology or without technology have been probably the biggest problems in humankind and any enterprise as well. So I guess that is also a big challenge because of why we are not seeing that scale up.

Vedica Kant

Romal, I’ll start with you kind of a couple of final questions before I open up to the audience. You know, we, you open up Twitter, there is always some kind of thread which is, I’m going to do this. And Claude has launched in PowerPoint, consultants are quaking in their shoes, you know, the skill set that you bring is seen to becoming like highly commodified, right? How scared are you of that disruption? That’s the first question. And how is AI also help, you know, forcing you or making you rethink your own pricing, your price points, et cetera, because, you know, our clients coming to you and saying, I can run this on ChatGPT, why do I need to pay you as much as I pay you?

So we just love your take on those two things.

Romal Shetty

Yeah, I think the first part is anything which is commoditized, I am scared, we are scared that that will completely go away. But can I do something? …So pretty cool. They saw a surge of demand, right, where people wanted to buy this stuff. But after some time, nobody was buying. So then they went in and figured out, you know, AI also did

Vedica Kant

On pricing.

Romal Shetty

On pricing. And the fact is that today, what I’m talking about the tax opinion, and we used to charge a particular sum of money, and we’ll charge a different sum of money. And people would say, hey, you know, you’re all cannibalizing stuff. But if I don’t cannibalize, or if I don’t do it, somebody else is going to do it anyway. so we’ve got to be open to it disruption is going to happen we can’t close our eyes but the fact is that also don’t get too hyped by every talk that the world will end tomorrow for all of you to the other extreme that nothing will happen I think the truth lies somewhere in between but keep looking at things to keep disrupting yourself and keep identifying newer sources of how your work can actually happen so I think that’s what it is

Vedica Kant

Can you just building on that how just given this point about pricing pressure etc how do you think about moving up the value chain are there other areas you think about going into and just when it comes to the model of consulting you’re seeing open AI, anthropic etc going saying we would want to we now need to implement our solutions we need to become consultants how much of a threat are you seeing from technology firms who are increasingly going down yeah yeah

Sanjeev Krishan

So maybe first thing first, I think this question is a bit unfair to consultants at large, right? Because I do believe, and we have seen multiple threats to consulting businesses in the past as well. I mean, forget AI. Over the last five years, every consulting firm, I’m sure yours included, would be saying that, okay, let me figure out, you know, how can I be more value -accurative to my client, right? What is the context of the client? What is the mindset of this client? I mean, are there generational issues? Are there succession issues? Are there technology issues? Are there business issues? Are there sustainability issues? Environmental issues? So on and so forth. And in a world which is so disrupted geopolitically and otherwise, supply chain, this, that, and the other, how do I either protect value or create value?

So from that standpoint, I think, you know, as I said, technology to me, or AI also, is a tool, is an enabler in that sense, right? It can help me contextualize better. It can help me simulate better. It can help me validate my assumptions a lot better. And in any case, over the last four to five years, as I said, most firms, most consulting, I’m not saying that there isn’t any time and material work for any of us. I’m sure there is. But let me also say that most of us actually have moved towards value accretion, value billing. And why would clients pay for something like that? I mean, that’s something which is getting commoditized.

In any case, I should feel threatened irrespective of AI. And today, in my mind, it is about how can I create value or defend my client’s value. So we ought to move up the value curve. A large part of billing for most consulting firms will come from the value that they create, whether it is simple cost optimization, whether it is some enterprise -wide transformation or segmental transformation, or indeed, you know, stuff like doing deals, raising money, and so on and so forth. So I do believe a lot of that has changed. The proportionality of that is possibly a little low on the lower side. It will possibly go up. So I think that’s the first thing.

To the second part of the question that you asked about, you know, about I think, you know, one has to acknowledge that we don’t need to do everything. I mean, if you think that we will be able to compete with a product firm, then I think we’re going down the wrong direction, in my mind at least. So certainly we want to work with a bunch of alliance partners, whether it could be, I mean, we were the first ones to partner with Harvey, for instance, which is OpenAI funded, and today a lot of our tax and legal work is actually done on the Harvey platform, for instance. So it is about how do we work with some of these disruptors or people who have taken pathways to the LLMs or so to speak.

And I do believe that, I mean, we recently are doing something with Anthropic now. So I think we will have to look at partnerships to be able to work with them. Again, as I said, the quantum of clients that we have globally is something which, you know, some of these disruptors will take ages to get to. And the context will require them to make very significant investments. So let me just round it off. I’ll have it once in the last point. you know people can say that there is disruption on tech and there is need for transformation but there is also disruption in trade yeah so today any tech transformation that you do let’s say on the supply chain side can you do it without a tax person involved can you do it without a trade specialist involved so it has to be trade and tech specialism which has to which has to come together to create value and that is why i don’t think that people who are writing obituary of the consulting model they’ll possibly have to wait so it’s a resilient model as you said has held its own for many years i’ll open up to the audience if we have any questions we can take a couple

Audience member 1:

yeah thank you hi hi i’m the founder of corral inc and my question to romol and sanju and my question and both redefining country power and people productivity. Right now, of course, USA and China are leading the race, but India is third. Where do you think that, you know, the next probably $100 billion to $500 billion company

Vedica Kant

I think the question was about whether you’ll see AI creating, let me paraphrase, but AI creating more abundance and societal impact. And are we going to see another, from India, a trillion, a $500 billion company? Or a billion dollars?

Sanjeev Krishan

Well, first of all, I’ll say that it better come from the U.S. Otherwise, all the amount, all the leverage and capital which has gone in the U.S. markets will come to nothing. And I’m sure a lot of people will lose a lot of money and the financial markets will get shaken up. But, you know, I think I do believe, I do believe that, you know, some of these, you know, I think it’s very early days yet. And people who are putting capital to work, I’m sure know what they’re doing. You know, I’m sure many of these things may not work out. And that’s the nature of venture capital business, for instance. Right. But clearly, you know, I think one thing which we can be certain about is that this is an irreversible trend.

I mean, AI is something which is going to stay with us. It is only going to get better. I mean, you know, today we are talking about, you know, AGI, for instance. Right. And that, you know, I’ve felt so far in my non -technical mind that, you know, technology can never compete with humans. But with AGI, it can, you know, it can go beyond humans as well. I mean, depending on what it does serve. So I do believe that there will be winners which will come through. I think it will possibly take nine, you know. getting the, for instance, there is no real TAM in my mind. You know, if I can be honest, there’s no real TAM in any market other than the US at this point in time.

So this will take time, but this is going to happen for sure. When it can come from India, you know, it’ll possibly take time. But the question really is that what will cause those to come? It will not necessarily come through, you know, the businesses that possibly work in the US. In my mind, we will have to find our own pathways. And I think this summit is a great opportunity to create those pathways. And then you know that our ability to you know, in some way scale those is very, very high. So I do believe that, you know, it’s going to be sequential. It’s going to happen. It may not be the most value -accredited thing that will come from India, but possibly we will be the first few ones to be

Vedica Kant

I think we had a few questions. I think the gentleman in the back had raised his hand, and then we can have a few here. But Leanne,

Audience member 2:

Hi, I’m Abhinav Saxena, consultant at Capacity Building Commission, Government of India. So we had a panel discussion, just thought of joining it, hearing from you. So I want to know how the GovTech space looks like, how the government consulting space looks like when we are seeing a lot of AI -based tools and AI -based interventions launched by the government. I would be happy to have your insights and share mine. I’ve recently had an entire state calibrated for an AI tool. It was a chaos, but somehow we managed. Yeah, your insights on this.

Romal Shetty

Yeah, so I mean, clearly I think it’s a big space for us. I think for all consulting firms, government is a big space where we’re all investing time and energy and we see very, very interesting propositions come out. I mean, to give you an example, one of the chief ministers told me that in the past, that, you know, Romal, I spent today on a road, on a stretch of road, which could be one kilometer. I could be spending 20 crores to 50 crores. Now, people tell me that maybe there’s topography, there’s demography, all of that stuff. And therefore, that’s the reason. But I’m not so sure. Can you help me assess through geospatial and AI?

Can you estimate, for example, why should what should it cost to build a road or to repair a road? I have a thousand crores loss. What is it that you can actually help me? So there are very different kinds of things coming from skilling to access to credit. Our MSME, for example, access to credit. I may get credit today at 8 percent. But if you take MSME, a lot of them may get 24 percent because they don’t have collaterals. But with the data that they have today, it may be much easier for financial institutions to give them at that 8 or 9 percent. So I think GovTech and in many places, and we clearly see India, for example, really pushing forward on that.

And a lot of our solutions that we’re doing here. probably going elsewhere as well. So clearly huge potential, huge opportunity.

Audience member 2:

We can expect your sample and collaboration with the giants for good sample and

Romal Shetty

Absolutely.

Audience member 3:

Namaste sir. I am a student. So my question is that what should be the effective strategy that students from rural areas or tier 3 cities should follow so that they can take maximum leverage of AI and what do you think will be the future of degree courses or our education system as everything is being restructured and possibly it may become obsolete. So what are your thoughts on it?

Romal Shetty

So as I said, I think the skills of the future are a little bit different. So really, like I said, you know, critical thinking, right? Judgment capabilities, working with machines, including humanoids, we will have. And of course, the ability to have access to various kinds of information that will help, especially in the rural areas. Do you have more practical based, but with AI actually helping you, you know, learn concepts better? Because I think the conceptual knowledge is more important than the rote, which used to happen. And then how do you sort of apply that? One important thing, and we talk about it in consulting firms, the ability to orchestrate. You know, I always say that, I mean, I don’t believe in palmistry, but, you know, for an example, we say that a good palmist reads one line, a better palmist maybe reads two lines, but a great palmist is able to read all the lines and make sense of that.

And in some sense, that is sort of the skill that you’ll have to start building, considering all kinds of, you know, things that impact your life.

Audience member 3:

So one more question is that. In fact, how humans and AI are…

Vedica Kant

Sorry, we have a lot of people who’ve raised their hands. I think we can just probably take a couple of questions. I think we had the lady here and then I think we can go to the gentleman in the back. Yeah. Please.

Audience member 4:

Hi, my name is Geeta. So, following on from the talent question and more so I come from a sort of a GCC background as well thinking of talent. The critical thinking, the power skills, so to say. Picking a grad or an undergrad or even for that matter an ACC or a CA with the current sort of rigor and the qualification and all of that and then transporting that talent into the newer world. It’s a bit of a tussle between the skills that are required today, the skills of tomorrow and how is it that the talent the student should be thinking and how is it you guys are thinking about it?

Sanjeev Krishan

So let me just say, you know, and I’m glad that you raised that question. I mean, at least in the last nine months, I’ve been advocating at whenever the opportunity presents itself, the need for us to do a bit of a rehaul of our education system. Many of my engineering friends tell me that 95 % of what they learned in BHU, for instance, or many of our engineering institutes remains the same as what is being taught 25 years back and what is being taught today is the same. I would have thought that maybe it should be 75%, maybe 80%. I mean, you know, the skill sets of what, as you said very rightly, what will be required tomorrow is going to be very, very different.

I mean, we see, we certainly see that many students today are taking psychology, for instance, and sociology, etc., apart from, and that actually goes to the point that Romul made earlier. So I think some of the skill sets are going to be different. But I must say that working with technology as opposed to, you know, working with technology, at technology, which is like coding as we were talking earlier, is going to be very very different. And I do believe that it requires us to teach a different curriculum in our schools also, not just colleges, schools also, and that is going to be a starting point. I do also want to mention, you know, in respect to the previous question which came in, I think you know, the whole AI piece is going to enable, you know, like the GCC industry, I’m sure, you know, like this question could easily be asked to the GCC industry.

This session could easily be for the GCC industry that how is GCC industry going to get disrupted by AI? I do believe that one of the things that we you know, as a nation and civil society should be focusing on is what does it do to entrepreneurship? Does it enable entrepreneurship at scale? Just as we are saying that UPI has enabled a certain amount of entrepreneurship, I think AI will be a huge enabler for entrepreneurship to the question that was previously asked and I suppose to the leverage that education can have for us.

Audience member 5:

yeah i am sudhakar gandhey former senior director american express bank and also build a technology company called access cadets technologies which is a hundred million dollar company in 10 years so i understand a little bit of finance and technology the question anybody can answer my question is lot of money has gone into ai a lot of coming whether it is google or microsoft and everybody raised billions of dollars and moved the market to trillions now one thing which is coming out we look at lot of wall street journals etc the money which is gone into these companies from there is gone to few companies to test it out ok so what is it possibly think this whole thing will be re -rated some of this the whole thing will be re -rated ok because first time google and microsoft both are going to debt market to raise hundred billion dollar which they never raised because they gone to debt because equity of money is going to be raised and they are going to raise hundred billion dollar almost dried up now So my question to any all the three of you who can answer this question, do you think this whole thing will be re -rated and you think some of these companies will go under the water or come down to half the value or one quarter of the value, then the real story starts.

That means this happening in next one year, two year, three years will be reworked into much longer time. So basically re -rating the whole thing, some of these companies going under the water. Thank you.

Romal Shetty

Whenever you work with any kind of disruptive technologies, there will be people who will go under the water, there will be people who sort of succeed and that’s a fact of life. So even in this cycle, I think you will have some companies that will do really well, some companies that may not do very well. I mean you see investments in data center for example, they are saying now you don’t need that much space, you probably need one third of this hall to have a good result. You have a pretty large data center. So I think that is possible. but as I said I always caution on doomsday scenario either ways this way or that way that everything will be everybody will make money and nobody will make money I think that’s not going to happen second is also as India we’ve got to figure out our own thing whether we focus more on the how to better use AI for different things whether society whether for government whether for our own enterprises not necessarily only build everything we do have people like Servam who built also phenomenal things at a lower cost but we’ve got to be very clear where we want to play and I think that is how we want to win I think that is what we should focus on but in these kind of things it happens we’ve also had that’s why if you look at the if you look at the SAP index you know last 25, 30, 40, 50 years those at 50 years back who are top companies don’t necessarily are there in the index now and that’s life that’s how evolution will always happen

Vedica Kant

I know we have a lot of questions but I think we all I don’t know if you have 5 minutes we’re going to take a short break Maybe we can take one more question because I think we also have to wrap up here quickly. I think we had one here and then one, the gentleman there. So I think we can do that as the last two questions.

Audience member 6:

So my question builds on something Romal said earlier in the session that your serviceability for SME clients is going to rise. But do you think SMEs are also better positioned? So from a demand perspective, is a lot of demand going to come from there because they are better positioned to leverage this neural network -driven AI because they don’t have to necessarily comply with data residency because most of these highly capable LLMs are housed not in India but elsewhere. And also this technology essentially is very probabilistic. So outcomes are going to be uncertain. And so the enterprise AI adaptation is mostly going to come from smaller firms, less regulated firms? Or do you think that’s not going to be much of a challenge because of the…

Romal Shetty

No, see, I think the… There could always be speed when it comes to smaller companies but doesn’t mean that the enterprises are not actually adopting. In fact, enterprises are spending a lot more. Regulated industries comes with its own thing because you have very strong regulated financial services, healthcare. They’ll be very careful of what they do. But I don’t think anybody is going to be left in this race or wants to be left out in this race. And everybody should be looking at what’s best for them. You don’t necessarily always need to go for LLMs that are… You can also go for open -sourced LLMs. So you don’t need to necessarily… And it’s a combination. I don’t think today there’s one that can solve all your problems.

There could be 10 different kinds of LLMs as well. And you have to be careful and choosy of what you want to do. The good part about the SMEs is they can leapfrog and not necessarily go through a big… cycle where they have to wait for 10 years to do things. And I think that it levels the playing ground a lot.

Audience member 7:

Hi, I am Piyush from Digivancy. My question is to Romul sir. As we talk about we can develop a campaign with an MNATS or something. So can we make a tool in terms of MarTech to find the right market for any of the new product line SKUs or for the SMEs because they do not have enough patience for like to do the research or some things even though big corporates as well.

Romal Shetty

Absolutely. So I mean if you do a sentiment analysis you can probably find markets where you think there is demand. I mean it’s like Google knows exactly when somebody is wanting a doctor, wanting something else. It knows actually right. How does it know? That’s the way it knows. So you can actually do some of these things and I do think especially in the SMEs side. the uberization of demand that is demand and supply we do it for taxis but really demand and supply for services or demand and supply for goods or whatever can be much much better because of this technology that we actually have

Vedica Kant

I just want to say thank you to everyone we had a really packed hall today thank you to our speakers for actually being very honest not all consulting leaders will necessarily be as honest also about how their consulting model is changing and shifting and the questions that they have to confront so thank you very much thank you thank you Thank you.

Romal Shetty

Speech speed

186 words per minute

Speech length

Macro‑economic concerns about AI investment valuation

Topics

Financial mechanisms | Artificial intelligence

Sanjeev Krishan

Speech speed

205 words per minute

Speech length

Talent development, education and future skills

Topics

Capacity development | Artificial intelligence

Vedica Kant

Speech speed

155 words per minute

Speech length

GovTech and broader societal impact

Topics

Social and economic development | Artificial intelligence

Audience member 2

Speech speed

157 words per minute

Speech length

111 words

Speech time

42 seconds

State‑level AI calibration challenges

Explanation

The audience member shares experience of calibrating an AI tool for an entire state, highlighting practical deployment issues.

Evidence

“I’ve recently had an entire state calibrated for an AI tool.” [22]. “So I want to know how the GovTech space looks like…” [55].

Major discussion point

GovTech and broader societal impact

Topics

Social and economic development | Artificial intelligence

Audience member 3

Speech speed

155 words per minute

Speech length

84 words

Speech time

32 seconds

Rural / tier‑3 student strategy for AI

Explanation

The audience member asks how students from less‑privileged backgrounds can leverage AI and what the future of degree programs will look like.

Evidence

“what should be the effective strategy that students from rural areas or tier 3 cities should follow so that they can take maximum leverage of AI and what do you think will be the future of degree courses or our education system as everything is being restructured and possibly it may become obsolete.” [128].

Major discussion point

Talent development, education and future skills

Topics

Capacity development | Closing all digital divides

Audience member 4

Speech speed

162 words per minute

Speech length

115 words

Speech time

42 seconds

Emphasis on critical thinking and power‑skills

Explanation

The audience member stresses that future talent needs strong critical thinking and power‑skills to work effectively with AI.

Evidence

“The critical thinking, the power skills, so to say.” [48]. “Picking a grad or an undergrad or even for that matter an ACC or a CA with the current sort of rigor and qualification… and then transporting that talent into the newer world.” [133].

Major discussion point

Talent development, education and future skills

Topics

Capacity development | Closing all digital divides

Audience member 5

Speech speed

202 words per minute

Speech length

277 words

Speech time

82 seconds

Re‑rating of AI‑focused companies

Explanation

The audience member wonders whether the massive AI investment wave will lead to a re‑rating of valuations, with some firms potentially failing.

Evidence

“So basically re‑rating the whole thing, some of these companies going under the water.” [109]. “What is it possibly think this whole thing will be re‑rated… some of these companies will go under the water or come down to half the value or one quarter of the value…” [151]. “It will possibly go up.” [152].

Major discussion point

Macro‑economic concerns about AI investment valuation

Topics

Financial mechanisms | Artificial intelligence

Audience member 6

Speech speed

132 words per minute

Speech length

130 words

Speech time

58 seconds

SME serviceability and AI advantage

Explanation

The audience member builds on Romal’s point, noting that SMEs can adopt AI more quickly and may become a major source of demand for AI services.

Evidence

“So outcomes are going to be uncertain.” [86]. “And so the enterprise AI adaptation is mostly going to come from smaller firms, less regulated firms?” [99]. “But do you think SMEs are also better positioned?” [100]. “Your serviceability for SME clients is going to rise.” [101]. “The good part about the SMEs is they can leapfrog and not necessarily go through a big… cycle where they have to wait for 10 years to do things.” [102]. “So you can actually do some of these things and I do think especially in the SMEs side.” [104].

Major discussion point

SME market opportunities enabled by AI

Topics

The digital economy | Artificial intelligence

Open‑source LLMs vs. proprietary models

Explanation

The audience member points out that enterprises can consider open‑source large language models to avoid data‑residency constraints.

Evidence

“You don’t necessarily always need to go for LLMs that are… You can also go for open‑sourced LLMs.” [97]. “And so from a demand perspective, is a lot of demand going to come from there because they are better positioned to leverage this neural network‑driven AI because they don’t have to necessarily comply with data residency because most of these highly capable LLMs are housed not in India but elsewhere.” [117].

Major discussion point

Challenges of AI adoption in enterprises

Topics

Data governance | Artificial intelligence

Audience member 7

Speech speed

154 words per minute

Speech length

75 words

Speech time

29 seconds

MarTech tool for rapid market research

Explanation

The audience member proposes building an AI‑driven marketing technology tool that can quickly generate campaigns and identify demand for new SKUs, especially for SMEs.

Evidence

“So can we make a tool in terms of MarTech to find the right market for any of the new product line SKUs or for the SMEs because they do not have enough patience for like the research…” [141]. “So I mean if you do a sentiment analysis you can probably find markets where you think there is demand.” [142]. “I mean it’s like Google knows exactly when somebody is wanting a doctor, wanting something else.” [144].

Major discussion point

SME market opportunities enabled by AI

Topics

The digital economy | Artificial intelligence

Audience member 1

Speech speed

96 words per minute

Speech length

60 words

Speech time

37 seconds

No specific contribution captured

Explanation

The transcript does not contain a distinct comment from Audience member 1, so no direct evidence is available for this participant.

Major discussion point

N/A

Topics

N/A

Agreements

Artificial intelligence | The digital economy | Closing all digital divides

Education systems require fundamental overhaul to prepare for AI-driven future

Speakers

– Romal Shetty
– Sanjeev Krishan
– Audience member 3
– Audience member 4

Arguments

Future skills focus on critical thinking, judgment, and orchestration abilities rather than rote learning

Education system needs overhaul as 95% of engineering curriculum remains unchanged from 25 years ago despite technological advances

Students from rural areas and tier 3 cities need effective strategies to leverage AI and adapt to changing education systems

There’s a significant challenge in transitioning current graduates with traditional qualifications to meet future AI-era skill requirements

Summary

All speakers agree that current education systems are inadequate for the AI era, requiring shifts from rote learning to critical thinking, with particular challenges for students from disadvantaged backgrounds and those with traditional qualifications

Topics

Capacity development | Artificial intelligence | Social and economic development

Data governance and security concerns are major barriers to enterprise AI scaling

Speakers

– Romal Shetty
– Audience member 6

Arguments

Data governance and security concerns prevent scaling, with companies finding their IP appearing in public AI systems through vendor usage

SMEs may be better positioned for AI adoption due to fewer regulatory constraints and data residency requirements

Summary

Both speakers identify data security and governance as critical challenges, with Romal providing specific examples of IP leakage and the audience member noting that regulatory constraints may disadvantage larger enterprises compared to SMEs

Topics

Data governance | Building confidence and security in the use of ICTs | Artificial intelligence

AI democratizes access to capabilities previously available only to large organizations

Speakers

– Romal Shetty
– Sanjeev Krishan
– Audience member 7

Arguments

SMEs can benefit from AI-powered marketing campaigns and market research tools that were previously unavailable to them

AI will enable entrepreneurship at scale similar to how UPI enabled entrepreneurship, with significant potential for societal transformation

AI-powered MarTech tools could help SMEs and corporations identify optimal markets for new products and SKUs

Summary

All speakers agree that AI levels the playing field by giving smaller organizations access to sophisticated capabilities, with examples ranging from marketing campaigns to market research and entrepreneurship enablement

Topics

Closing all digital divides | Artificial intelligence | The digital economy

Social and economic development | Artificial intelligence | Information and communication technologies for development

Financial mechanisms | Artificial intelligence | The digital economy

Overall assessment

Summary

The speakers demonstrate strong consensus on AI’s transformative potential while acknowledging significant implementation challenges. Key areas of agreement include the need for business model transformation, workforce reskilling, educational reform, and the importance of change management. There’s also consensus that AI will democratize access to capabilities and create new market opportunities, particularly for underserved segments like MSMEs.

Consensus level

High level of consensus on strategic direction and challenges, with speakers showing realistic optimism about AI’s potential while being candid about implementation difficulties. The agreement spans both opportunities (new markets, democratization) and challenges (change management, data governance, financial sustainability), suggesting a mature understanding of AI’s implications for consulting and broader society.

Differences

Different viewpoints

Certainty about future pyramid structure in consulting firms

Speakers

– Romal Shetty
– Sanjeev Krishan

Arguments

The traditional pyramid structure will reshape with middle management shrinking in some areas while creating new opportunities in underserved markets like MSMEs

The type of people hired will be very different, with associates potentially doing manager-level work, requiring different workforce planning

Summary

Romal provides specific predictions about how the pyramid will change (middle shrinking, new MSME opportunities), while Sanjeev admits uncertainty about the pyramid’s future shape, stating ‘I don’t know the answer to the pyramid question’ but focuses on changing skill requirements

Topics

Artificial intelligence | The digital economy | Capacity development

Approach to AI disruption and market positioning

Speakers

– Romal Shetty
– Sanjeev Krishan

Arguments

Commoditized work will disappear, but firms must disrupt themselves rather than wait for others to do it, while being realistic about timeline

AI acts as a utility that firms must embrace, with PwC committing nearly a billion dollars to AI platforms and upskilling

Summary

Romal emphasizes proactive self-disruption and cannibalization of existing services, while Sanjeev focuses on AI as an enabling utility that enhances existing capabilities rather than disrupting them

Topics

Artificial intelligence | The digital economy | The enabling environment for digital development

Unexpected differences

Timeline and urgency of AI transformation

Speakers

– Romal Shetty
– Sanjeev Krishan

Arguments

AI enables reimagining business models, inverting the traditional pyramid from 1:10 to 10:1 client-to-person ratio

All personnel should have access to AI tools like ‘Chat PwC’ to create efficiency and inspire client solutions

Explanation

While both are AI advocates, Romal presents more radical transformation scenarios (complete business model inversion) while Sanjeev focuses on gradual integration and employee empowerment. This suggests different views on transformation pace and scope

Topics

Artificial intelligence | The digital economy | Capacity development

Overall assessment

Summary

The speakers show broad alignment on AI’s transformative potential but differ on implementation approaches, timeline expectations, and specific strategic responses. Key disagreements center on workforce restructuring certainty, disruption vs. enhancement strategies, and the balance between technical vs. human factors in AI adoption challenges.

Disagreement level

Topics

Artificial intelligence | The enabling environment for digital development | Financial mechanisms

Social and economic development | Artificial intelligence | Information and communication technologies for development

Takeaways

Key takeaways

AI is fundamentally reshaping consulting business models, enabling firms to invert the traditional pyramid structure from 1:10 to 10:1 client-to-person ratios through human-AI collaboration

The consulting workforce will require different skills focused on critical thinking, judgment, empathy, and orchestration abilities rather than routine data processing tasks

Enterprise AI adoption faces significant challenges with only 12% of corporations achieving both top-line and bottom-line benefits despite substantial investments

Change management and integration are the primary obstacles to AI scaling in enterprises, not the technology itself

Data governance, security concerns, and potential token pricing shocks are major barriers preventing pilot projects from reaching production scale

Consulting firms must embrace partnerships with AI disruptors rather than compete directly, while focusing on value-based billing over commoditized work

AI will democratize access to previously expensive services, particularly benefiting MSMEs and enabling new market opportunities for consulting firms

The education system requires fundamental overhaul to prepare students for AI-enabled work environments, emphasizing conceptual learning over rote memorization

AI presents significant opportunities in GovTech applications and will enable entrepreneurship at scale similar to UPI’s impact

Market re-rating in AI investments is inevitable with some companies succeeding while others fail, which is natural in disruptive technology cycles

Resolutions and action items

Consulting firms should continue investing in AI platforms and upskilling their workforce to remain competitive

Firms need to develop partnerships with AI technology providers like Harvey and Anthropic rather than building everything in-house

Organizations should focus on change management and integration strategies when implementing AI solutions

Educational institutions should revise curricula to include AI collaboration skills and reduce outdated technical content

India should identify its own AI pathways and focus on leveraging its scaling capabilities rather than directly competing with US markets

Consulting firms should accelerate their transition to value-based billing models as commoditized work becomes automated

Unresolved issues

The exact shape and structure of the future consulting workforce pyramid remains unclear and will likely vary by sector and competency

How to effectively manage data governance and IP protection when using AI tools across vendor networks

The timeline and extent of market re-rating in AI investments and which companies will survive the transition

How to address the potential ‘bill shock’ when AI token pricing moves from subsidized to market rates

The specific curriculum changes needed in educational institutions to prepare students for AI-enabled careers

How smaller firms and SMEs will handle complex AI implementations without extensive technical resources

The long-term competitive dynamics between traditional consulting firms and technology companies entering the consulting space

Suggested compromises

Use a combination of different LLMs and open-source models rather than relying on single proprietary solutions to balance capability and cost

Implement a gradual transition approach where AI augments human work rather than immediately replacing entire job functions

Focus on human-in-the-loop AI implementations to maintain quality control while gaining efficiency benefits

Develop hybrid business models that serve both traditional enterprise clients and new MSME markets with different service approaches

This comment is insightful because it identifies the fundamental paradox of technological adoption – humans create technologies they then resist adopting. It shifts the focus from technical capabilities to human psychology and organizational behavior, suggesting that the real barriers to AI success are social rather than technical.

Impact

This observation reframed the entire discussion about AI implementation challenges. Instead of focusing on technical limitations, it directed attention to change management, which became a central theme for explaining why AI pilots don’t scale and why enterprise adoption remains low despite significant investments.

Overall assessment

Explanation

Krishan suggested India needs to find its own pathways for AI success but didn’t specify what these pathways should be

Disclaimer: This is not an official session record. DiploAI generates these resources from audiovisual recordings, and they are presented as-is, including potential errors. Due to logistical challenges, such as discrepancies in audio/video or transcripts, names may be misspelled. We strive for accuracy to the best of our ability.

AI Transformation in Practice_ Insights from India’s Consulting Leaders

Summary

Keypoints

Major Discussion Points:

Overall Purpose:

Overall Tone:

Speakers

AI as a Fundamental Business Model Disruptor

Internal AI Implementation and Workforce Transformation

Enterprise AI Adoption Challenges and Market Reality

Addressing Disruption and Competitive Threats

Educational System Reform and Future Skills

Government Technology and Societal Impact

Market Dynamics and Financial Sustainability

Future Pathways and Strategic Recommendations

Business‑model inversion

Explanation

Evidence

Major discussion point

Topics

Audit‑confirmation automation saves 60 000 hours

Explanation

Evidence

Major discussion point

Topics

Rapid simulation for manufacturing

Explanation

Evidence

Major discussion point

Topics

GovTech – MSME credit access

Explanation

Evidence

Major discussion point

Topics

Middle‑layer shrinkage in consulting pyramid

Explanation

Evidence

Major discussion point

Topics

Data‑governance and IP risks

Explanation

Evidence

Major discussion point

Topics

Pricing pressure on tax opinions

Explanation

Evidence

Major discussion point

Topics

SME leap‑frog potential

Explanation

Evidence

Major discussion point

Topics

Macro‑economic AI investment cycles

Explanation

Evidence

Major discussion point

Topics

Enterprise‑wide “Chat PwC” platform

Explanation

Evidence

Major discussion point

Topics

Navigate Tax Hub – AI‑driven tax tool

Explanation

Evidence

Major discussion point

Topics

Manager tasks shift to associates

Explanation

Evidence

Major discussion point

Topics

Change‑management and integration hurdles

Explanation

Evidence

Major discussion point

Topics

Audit‑confirmation automation saves 60 000 hours

Enterprise‑wide “Chat PwC” platform