AI Is Ready — People Aren’t

Menrv.AI is positioned as the world’s first AI-as-a-Service platform with over 60 AI products. What inspired you to build it, and how did your experience at companies like Disney and Sony shape that vision?

Thank you for having me. It’s always a pleasure to be back at XLRI — I studied here, did both my MBA and PhD, and have been teaching here for years. I even received the Distinguished Alumni Award last year, which was a proud moment.

Coming to your question — yes, Menrv.AI is one of the world’s first AI-as-a-Service platforms. The inspiration really stems from my professional journey. Over the past 10-15 years, I’ve worked with leading Fortune 100 companies like Disney, Sony and Target, as well as major startups such as Grab in Singapore and a large e-grocery venture in Indonesia.

Back when I was doing my MBA, data science and AI weren’t even part of the vocabulary. Everyone wanted to go into marketing, finance, or strategy. But over time, I saw the landscape shift — first in Silicon Valley and gradually across the globe. We moved from the mobile and internet era to big data, which drove the need to analyse information and spot patterns. That gave rise to analytics, which evolved into data science, and eventually AI.

I was fortunate to be in industries where AI adoption was particularly rapid — e-commerce, tech, and OTT platforms like Disney+ and SonyLIV. As part of my roles, I helped develop various AI products. At Sony, I headed data science for Sony R&D India and helped establish its first R&D centre in the country. We built innovative algorithms — for example, the recommendation engines you see on platforms like Hotstar or Netflix. Those were the kinds of tools we were working on.

What struck me, though, was that while these models delivered significant impact — generating millions in additional revenue and profit — they were limited to the company that built them. A recommendation system built for Disney, for instance, would never be shared with anyone else.

Yet there are countless retail, e-commerce, and logistics firms around the world that could benefit from such tools. Many lack awareness, talent, or resources. That’s what sparked the idea for Menrv.AI — to shift from building AI solutions for one company at a time, to offering them as a service. Much like Software-as-a-Service, the idea was to give companies access to powerful AI tools at a fraction of the cost, with no lead time.

That was the genesis of Menrv.AI. We’ll probably get into more details as we go.

So you’ve essentially transformed these AI products into ready-to-deploy models. Is that right?

Not quite. Our approach is a bit different. Everyone now agrees that AI is useful — that’s not in question anymore. But the real challenge is accessibility. How do people actually use it? Are they expected to learn programming, calculus, statistics, and so on? That used to be the case five or ten years ago — if you wanted to benefit from AI, you had to master coding and advanced maths.

But if that’s the requirement, we’re not really talking about democratising AI, are we? Because we can’t expect everyone in the world to become data scientists. Yet everyone should be able to benefit from it.

To solve that, I often draw parallels with two major tech innovations. First, early computers had command-line interfaces — black-and-white screens where you had to type every instruction. They weren’t accessible to the average person. But once graphical user interfaces came along — Windows, for instance — and you could just point and click with a mouse, computers became mainstream. The same thing happened with smartphones. Before the iPhone, phones were clunky and hard to use. Now, even a five-year-old can navigate apps and take pictures.

That’s exactly the vision behind Menrv.AI — to make AI as easy to use as a smartphone app. Because the moment you say, “deploy this code,” you lose most people.

When ChatGPT launched in late 2022, it was a great validation of our approach — but it was limited to chat-based applications. The scope of AI is much broader, especially for business use cases.

All our AI products are designed to be just as intuitive, as easy as using an app. If someone has access to the platform, all they need to do is log in, input their question or data, and that’s it. The rest — from front end to backend, cloud infrastructure to execution — is fully integrated and handled by the system.

So, yes, users only need their data and their problem statement. The platform does the rest. 

 

AI is often framed in technical terms — and when I said “deploy AI”, I meant exactly that. But you’ve built products for sectors like retail and finance, where users may not be technically inclined. Were there any surprising insights that led you to change your approach — either in the product itself or how you marketed it?

Yes, we had several pivot moments, and I think that’s crucial for any startup. In the early days, we were developing products while also taking on client projects and offering consulting and training, especially with top business schools.

One early client was a retail fashion company in South America. They wanted a recommendation engine, so we built a customised model for them over three or four months — a significant investment in time and effort, with no guarantee of success. Fortunately, it worked well, and the client was happy.

But then came the pricing conversation. We calculated our costs — five people over several months — and quoted what we thought was a fair price based on industry norms. But they pushed back, saying it was too expensive.

That moment forced a rethink. We realised this model — building bespoke solutions for individual clients — was flawed. It’s how most software and AI solutions are still developed: either in-house or outsourced to a vendor. But the issue is scalability. A highly customised product may not be useful to anyone else, and if the client walks away, you’re left with a sunk cost.

That experience was a turning point. We decided we didn’t want to be a consulting or training company, or even do project-based work. We would focus entirely on being a product company — even though it was the riskier option in the short term.

But long term, it made more sense. If we built products for marketing, sales, finance or retail, we could offer them to any number of customers. If one wasn’t interested, someone else might be.

It also made pricing more accessible. Let’s say it costs us $100,000 to build a product — that sounds steep to a single client. But if we sell it to 100 clients, it brings the cost down to $1,000 each — which is not only reasonable, but scalable. And that’s how many of the major AI firms operate: they invest heavily upfront, knowing they’ll scale the solution.

So the insight didn’t come from AI itself, but from our experience building and trying to sell it. That shift — from a service-led mindset to a product-first strategy — has defined how we’ve operated over the past two years. We’re still refining our product-market fit, but that strategic decision changed everything.

 

You’ve spoken about what worked in your favour. Can you share a failure — personal or professional — that turned into a pivotal learning experience?

As a startup, I don’t really see failure as a negative. Even in life, my belief is: you don’t need to succeed every time. You might fail 99 times — you just need that one success to break through.

And in a startup, it’s a daily process. Every day brings new challenges, setbacks, surprises. That’s what makes it exciting.

One of our key learning moments — you could call it a failure — came after we launched our product last year. We gave broad access, and the initial interest was promising. People were using the platform, so we assumed they’d convert into paying customers. But conversion turned out to be much harder.

Some said the pricing was high, others wanted tweaks or custom features. Many simply didn’t feel ready to adopt it. Think about it: even if 10 out of 100 users convert, that’s considered a good rate. But it also means 90 per cent didn’t pay. So from one angle, that looks like a 90 per cent failure rate.

But this isn’t unique to us. Take OpenAI — they reportedly have over 200 million users, but only 2-3 per cent pay for subscriptions. So by that logic, they’re “failing” 97 per cent of the time too. But in reality, even the free users are part of a learning process. The product evolves based on usage, feedback and behaviour.

We’ve taken the same approach. If a customer gives negative feedback, I actually welcome it. That tells us exactly where to improve. The worst is silence — when someone uses the product but doesn’t say anything. You learn nothing from that.

In fact, just yesterday I was speaking with a professor who’s been using our platform. I asked for her thoughts and she said, “It’s a great report, but one section isn’t working properly.” I appreciated that honesty — that kind of feedback is how we grow.

Even internally, we try to foster a culture where mistakes are okay. We encourage people to take risks. The only rule is: learn from it and don’t repeat the same mistake.

 

Since we’re talking about the future of AI, if you had to predict one groundbreaking application that could redefine how we live and work over the next decade, what would it be?

I’d say that shift is already under way — not just because of ChatGPT, but through many other AI tools, including what we’re building at Menrv.AI. The technology exists today and has the potential to permanently change how we work. The real issue isn’t availability — it’s awareness and behaviour.

Take MBA graduates, for example. When they start working, much of their day involves reading documents, writing reports or emails, and analysing data in Excel — whether they’re in consulting, marketing, finance or HR. Or consider this interview: you’ve likely spent considerable time preparing your questions. Now, all of that — reading, writing, analysing, content creation — can already be done faster and more efficiently with AI. You know that, and so do many others.

You can upload a document and ask AI to summarise or answer questions about it. You can ask it to draft an article. You can even feed it my profile and say, “I have to interview this person — suggest questions for a one-hour conversation.” You can always refine its output, but the point is: it can already do these things. Whether it’s language tasks, maths, data, audio, images or video — AI is capable.

So why aren’t we seeing mass adoption yet? From our experience, it’s not due to technical limitations — the tools exist. The problem is twofold. First, awareness is still limited. Even if ChatGPT has 200 million users, that’s just a tiny fraction of the world’s 8 billion people. Most have never heard of these tools, let alone used them effectively.

Second, even among those who are aware, habits are hard to change. That’s what we’ve observed repeatedly. When we demo our tools, people are excited. They use it actively at first. But over time, usage sometimes drops off — not because the product isn’t useful, but because people revert to familiar ways of working.

We’ve been reading documents, writing notes, or analysing spreadsheets manually for decades. Even small things — like preferring a printed question list, or reading a physical newspaper — are habits built over years. And habits are hard to break.

There’s also a certain satisfaction in doing things yourself — reading a newspaper, writing with a pen, or even typing something out. It gives a sense of depth or engagement.

True, but the reality is — you no longer need to write or read in the traditional sense. You can simply talk to an AI and get your answers. You can ask it to write for you. These tools already exist, and many of us — including your audience — are aware of them.

 

But the big question is: will people actually change their habits? And the answer, I think, is that it will take time — especially for those who’ve spent decades doing things a certain way.

It’s similar to what happened with smartphones. When they became mainstream around 2008, the younger generation embraced them immediately. But for older generations — like our parents or grandparents — it was a much harder transition. My father, for instance, was able to adapt from no phone to landline to feature phone. But once smartphones and tablets arrived, even simple things like answering a WhatsApp video call became a struggle — not because he wasn’t smart, but because he had no prior exposure to that interface or behaviour.

It’s the same with AI. Most people over 30 — even if they find it exciting and understand the efficiency gains — haven’t spent the last two decades using it. They’re conditioned to read, write and analyse manually. Changing that won’t happen overnight.

But the younger generation is a different story. Schoolchildren today are already using ChatGPT to help with homework or learning tasks. For them, this will be second nature — just another tool in daily life.

So yes, the technology is already here — and it will change how we live and work. But will that transformation happen immediately? I don’t think so.

Psychological shifts and habit changes take time. But five to ten years from now, we’ll likely be in a very different world. Gen Z will be joining the workforce and they’ll find it strange if you aren’t using AI. They’ll say, “Wait — you actually wrote this yourself? You analysed data manually?”

That mindset shift is what will ultimately drive the change.

 

We spoke earlier about Gen Z entering the workforce. I’m curious: how do you see workplace dynamics evolving because of this shift? And how should leaders adapt?

It’s definitely evolving — though I’m not sure if it’s purely generational or more about mindset and context. Based on my experience working with younger team members across companies, I do see differences — in behaviour, focus, and expectations.

But I wouldn’t generalise too much. Some of it may be generational — but a lot is individual. For instance, I’ve always been quite a serious person. When I was at XLRI, I was all about study and performance. I didn’t speak much in my first year, but worked hard to top the second year. That drive has always been a part of me.

That said, we maintain a relaxed culture in the company. If your work is good, take leave, enjoy your life — we don’t track log-in times or locations.

Still, with some younger employees — Gen Z or otherwise — I’ve noticed traits like restlessness, impulsiveness, or lack of sustained focus. And I don’t blame them entirely. Today, we’re all living with constant digital distractions — notifications, calls, emails, messages. Even preparing for this interview, I was getting pinged from every direction.

Now, for my role as CEO, that level of context-switching is manageable. I’m supposed to be interacting with people, switching gears. But if you’re an engineer, for example, and your job is to write high-quality code efficiently, that level of distraction becomes a real problem.

And then there’s social media. Some of the content out there — especially from influencers — can distort reality. You see someone claiming to make a million dollars a day just by posting on LinkedIn, with no background or credentials, and it creates this illusion of easy success. That’s dangerous. It can erode respect for hard work, patience, and long-term thinking.

Again, I’m not saying everyone is like that. I’ve met many young people who are focused, dedicated, and driven. But I do think the environment they’re growing up in — saturated with content, noise and comparisons — makes staying focused that much harder.

So ironically, the younger generation may need to work even harder at avoiding distraction. They need to set clear goals and follow through — and that requires discipline.

 

What, in your view, can companies do to support this transition?

Companies definitely need to adapt — because this is the future workforce. Gen Z is incredibly talented, especially on the tech side, and they bring a lot of energy and ideas. But they also have different expectations, and we’ve found that some adjustments help.

I’ll share what’s worked for us at Menrv.AI. We’ve built a flexible remote work culture. We have offices in Dubai and Bengaluru, but employees can work from home — or anywhere, really. We don’t track log-in or log-out times. There’s no fixed schedule. Each team just has one daily check-in, so we stay aligned and know who’s doing what.

The focus is purely on outcomes. Each person is given their tasks and timelines — as long as they deliver, we’re happy. This works especially well for younger employees, who often value flexibility and work-life balance. Some might not feel productive in the morning, but they’re more active in the evening — and that’s fine, as long as the work gets done.

We’ve also noticed many team members have personal passions and hobbies, which we actively encourage. One of our employees is a passionate photographer — she’ll occasionally take a couple of days off for a photography trip, and we support that. Another loves trekking — he recently went on a four-day trek. Someone else is off exploring the mountains with friends. We’re completely okay with these breaks, provided they’re honest and committed to delivering on their return.

This flexibility helps create a sense of trust and ownership. I’ve seen people stay with us longer because they genuinely value this approach.

Of course, it’s not perfect. Some individuals do take advantage — they’re distracted or unmotivated, and they don’t use the flexibility productively. But in a small team, it’s easy to spot who’s delivering and who’s not. I’m not sure if this model would scale in a large corporate setup — so I wouldn’t prescribe it universally. But for us, it works.

What I’ve also observed is that many young professionals today don’t prioritise just money or status. They value meaningful roles, learning opportunities, and good culture. They’re not looking to burn themselves out for 80 hours a week — they want to grow, learn and maintain balance. And I think that’s a good thing.