AI Creates Value Only When Embedded in the Business

You’ve spent over 18 years in AI and analytics. Could you take us back to the beginning? What first sparked your interest in this field?
This question takes me all the way back to my engineering days. We were immersed in mathematics, programming, and software
design, and I chose an elective in data mining that proved transformative. The course introduced me to neural networks, regression, and the fascinating idea that historical data could predict future outcomes. That insight sparked a curiosity that kept pulling me further into the world of data. Interestingly, I didn’t begin my career in analytics. I started with a banking product company, but because that early exposure had already shown me the potential of this field, I returned to it soon after. What kept me hooked over the years was the multidimensional nature ofAI and analytics — the unique intersection of data, technology, and domain expertise. You find yourself working at a point where deep technical methods meet complex business objectives, allowing you to solve some of the most strategic problems organisations face.
That blend of intellectual challenge, impact, and continuous learning is what has kept me motivated for nearly two decades.
You’ve led high-performing teams at Tiger Analytics. What’s your secret to building teams that thrive in such a fast-changing space like AI?
It has been an evolving learning journey. Building high-performing teams has two equally important dimensions: setting them up correctly from the outset and ensuring they continue to excel as they scale. To do this effectively, leaders must continually invest in their own growth, staying aware of shifting market dynamics, evolving technological landscapes, changing client expectations, and the aspirations of the people they lead. One of the biggest lessons I’ve learned over the years is the importance of understanding individuals deeply. Every person comes with unique motivations, strengths, and aspirations, and when teams grow, it becomes easy to group people and overlook these nuances. But the magic happens when you recognise each individual for who they are and align them with opportunities that resonate with them. Equally important is creating a culture of openness and psychological safety. Some of the best ideas I’ve seen have come from the youngest members of a team, and that only happens when people know they can speak freely without fear of judgment or failure. Another critical aspect is the foundation stage of team building. I often advise my managers to dedicate nearly 80 per cent of their oversight time to the first month or so, as this is when trust is formed, conflicts emerge, interpersonal dynamics take shape, and team synergy is established. If the foundation is correct, the team becomes a well-run engine that can sustain high performance. I like to think of teams as the wheels of a vehicle; unless those wheels are strong and well-maintained, the car won’t move forward, no matter
how powerful the engine is. At Tiger Analytics, we’ve also embedded continuous learning and openness into our culture through initiatives like Team Health Checks and the Tiger Listens forum. These platforms reinforce the idea that building a strong team is not a one-time effort but an ongoing process that requires attention, empathy, and adaptability. When you combine individualised attention, a psychologically safe environment, and a strong foundation, you create teams that not only perform but thrive in a fast-moving, constantly evolving field like AI.
You’ve worked with Fortune 500 companies across banking, financial services, and insurance. How have you seen AI truly transform these industries?
Looking back over the last 18 years, the transformation has been remarkable, almost evolutionary, unfolding in distinct waves.
In the early days, BFSI (banking, financial services and insurance) was one of the first sectors to adopt traditional data and
analytics. The focus was on credit scoring engines, fraud detection, building Excel-based dashboards, and generating insights. AI at
that time was very targeted and narrow.
Today, things look entirely different. AI has moved from being a “nice-to-have” to becoming a board-level priority.

Five years ago, we used to pitch AI use cases to CXOs, convincing them of the potential impact. Today, the conversation has
reversed. Clients already know which use cases they want to implement. They come to us not for ideas, but for execution,
scaling, and co-creation.
We now help organisations build:
Data and AI strategies.
Roadmaps that prioritise 100-plus use cases.
Enterprise-wide AI foundations.
A recent example is a leading bank that we helped transform its pricing strategy using data and AI, unlocking over $50 million
in potential incremental revenue.
This is a clear illustration of where the industry is headed: AI is no longer a support function; it is becoming core to business
strategy, competitive advantage, and revenue growth.
Even today, many chief executives hesitate about adoptingAI. What do you think are the biggest concerns driving these conversations?
If you had asked me this question two or three years ago, the hesitation would have been far stronger. Back then, AI wasn’t even part
of most boardroom agendas. However, with the rise of large language models and generative AI, this has undergone a dramatic
change. Today, every CXO recognises that AI is not just helpful; it is essential. Yet, despite this shift in mindset, specific concerns
continue to surface, some of which are grounded in reality and others stem from misconceptions.
One of the biggest myths I encounter is the belief that AI implementation must be a massive, enterprise-wide undertaking, a sort
of “big bang” transformation that requires overhauling processes, infrastructure, and entire organisational systems. Many leaders
wonder whether they are even ready for such an undertaking or whether it would be too complex to manage. Our advice is always
that meaningful impact does not come from launching AI everywhere at once. Instead, it comes from identifying a few high-impact
use cases, executing them well within a focused functional area, and gradually building maturity from there. AI does not need grand
gestures; it requires clarity, structure, and focus.
The second concern is far more valid and revolves around regulation, compliance, and governance. Industries like banking
and financial services are heavily regulated. As global AI regulations continue to evolve, with frameworks like the EU AI Act and
US guidelines emerging but not yet fully standardised, CXOs naturally proceed with caution. They worry about explainability,
fairness, transparency, model drift, and potential biases. They want to ensure that whatever AI-driven decisions they deploy are
compliant, ethical, and safe for customers. These are genuine concerns and entirely justified.
A significant portion of our work today involves helping organisations navigate this exact space — building responsible AI
frameworks, establishing governance practices, strengthening documentation, and ensuring that AI systems can be adopted
safely and at scale. This thoughtful, structured approach is what ultimately fosters trust and enables organisations to adopt AI
with confidence.
With evolving global regulations, such as the EU AI Act and new US guidelines, how do you see regulatory shifts impacting
AI adoption? What concerns do CXOs raise most often?
Regulation is one of the most active areas of discussion right now. The EU has introduced the AI Act, the US has released its
guidelines, and many countries are drafting their own frameworks. It’s an evolving space, and understandably, leaders want to know:
How will this impact our processes?
Will we remain compliant?
How do we manage fairness and transparency?
Will this affect our customers or our risk posture?
These are valid questions, especially in sectors such as banking and financial services, which are highly regulated. This is
not a challenge one group can solve alone. Industry, AI solution providers, and academia will need to collaborate. Over time, the
proper guardrails will emerge. However, today, it remains one of the most significant areas of concern.
The second concern we often hear is around impact timelines. CXOs fall into two extremes:
“AI will take years before it adds value.”
“AI can solve everything instantly.”
Neither is true. The reality lies somewhere in the middle.
When a use case is thoughtfully designed, aligned to business goals, and has clarity on how it integrates into workflows, value can
be delivered in as little as three to six months.
So the conversation today is shifting from: “Should we adopt AI?” to “How quickly and responsibly can we scale it?”
Tiger Analytics is known for delivering full-stack AI solutions. How do you ensure that these solutions are not only technically
strong but also practical and impactful for businesses?
This is a great question, and it reflects one of the biggest lessons I’ve learned over the past 17 years: the real impact of AI
never comes from the model itself; it comes from the business problem it solves. I’ve seen incredibly sophisticated models
end up gathering dust simply because they weren’t seamlessly integrated into business workflows or failed to influence
decision-making on the ground. AI creates value only when it becomes embedded in the day-to-day realities of the business.
At Tiger Analytics, we follow a business-persona-first design approach, something we summarise as “whiteboard before
keyboard.” Before writing a single line of code, we spend time deeply understanding the problem we’re solving, the end user
who will rely on it, the decisions that need to be influenced, the internal sponsor championing the investment, and the process
changes required for adoption. We also evaluate whether the client is organisationally and technically ready, because the
effectiveness of AI depends as much on readiness as on design.
Often, AI solutions span multiple stakeholders. For example, a cross-sell marketing engine requires alignment across data
teams, IT teams responsible for integrations, digital vendors accountable for deployment, and risk teams that ensure
compliance, especially in regulated industries. This is why we map out the business personas, their priorities, KPIs, and
action levers before we ever touch the data. Only after the business context is fully understood do we proceed to identify the
right data sources, validate data quality, and select the appropriate analytical techniques.
This structured, context-driven approach ensures that our solutions are technically robust, operationally practical, and capable of
driving real, measurable impact. It’s what transforms AI from an impressive prototype into a powerful business engine.
Tiger Analytics has developed plug-and-play AI accelerators. How do these make AI adoption simpler and more accessible for
organisations?

Plug-and-play accelerators have genuinely been a game changer for us, and analysts like Everest, ISG, and others have
recognised that. These accelerators are part of our broader innovation agenda and stem from our open IP-led approach, where
we believe in sharing our IP with clients.
Think of these accelerators as pre-built, domain-wrapped modules, like Lego blocks. They’re fully customisable, because
every client is different.
Traditionally, building AI solutions meant starting from scratch: writing code, testing, and customising for each client’s data
and workflows.
This is a long process. Over the years, the most significant request from clients has become the need for speed to value.
Everyone wants to realise impact faster.
That’s where accelerators make a difference. By packaging our learnings and automation into reusable modules, we can
deploy them quickly with minor customisations. This leads to a 40-60 per cent acceleration in time-to-value.
For example, a typical market-mix modelling solution typically takes three to six months to build. Using our accelerators, we can
deliver an MVP in six to eight weeks.
We currently have over 30 accelerators across various industries and functions, all tech-agnostic and field-tested. We fully
share the IP with our clients. They can use it, extend it, and retain it even if the engagement ends.
So plug-and-play accelerators have become a critical part of how we ensure speed, reliability, and scalability in
client engagements.
When helping companies build analytics Centres of Excellence (COEs), what are the biggest challenges you see? How do you
overcome them?
I’ve had the opportunity to set up and scale COEs for more than 15 years across banks, insurers, and financial services firms,
and I can say that building a successful COE is truly a combination of art and science. COEs are expected to be engines of both
innovation and delivery, and balancing these two expectations is often the most challenging part. Broadly, the challenges fall into
three interconnected areas.
The first is organisation-wide alignment. Within most client organisations, different functions operate with very different
priorities: marketing focuses on growth, finance is oriented toward cost optimisation, IT drives its own technology roadmap, and
operations focuses on efficiency. If these groups are not aligned, the COE struggles to gain traction. That’s why we begin by
ensuring that the leadership team is fully aligned and by establishing three to five clear business outcomes the COE must
achieve over the next few years. These outcomes become the North Star for everything we do, ensuring consistency and
direction.
The second challenge is talent. The competition for strong analytics and AI talent is intense, and while hiring is difficult, retaining the
right people is equally critical. At Tiger, we take this very seriously. Our attrition rate is under 10 per cent, well below industry standards.
We focus on building learning pathways, offering clear role definitions, creating skill-based career progression, and maintaining
continuous engagement. A COE ultimately reflects the strength of the people who make it, so talent becomes the foundation on which
everything else rests.
The third challenge lies in striking a balance between innovation and high-quality business-as-usual delivery. A COE is expected to

introduce new ideas, build new capabilities, and drive transformation, but at the same time, organisations depend on it for reliable, high-
quality, everyday analytical output. Managing this dual mandate, ensuring stability while pushing the boundaries of what’s possible, is

one of the most challenging but most crucial aspects of running a COE.
In my experience, these three factors — alignment, talent, and balance — decide whether a COE becomes a true strategic
asset for the organisation or simply another function within it. The most successful COEs are the ones that get all three right.
You work closely with both talent and academia. How do you envision this partnership shaping the future of theAI workforce?
Over the years, I’ve interacted with more than 2,000 students across B-schools and universities in India, and one thing stands out
unmistakably. There is a gap between what the industry needs and what academia currently provides, and both sides are aware of it.
The pace at which AI and analytics are evolving is simply much faster than what traditional academic structures can keep up
with. Even for those of us working in the field every day, staying current requires constant learning and agility.
To bridge this gap, we need much deeper collaboration between academia and industry. One area where this partnership can
make a significant impact is curriculum development. This idea has been discussed for years, but the implementation on the ground
has been limited. AI curricula need to evolve almost in real time, and that can only happen when industry practitioners actively
contribute to designing and updating them. The second opportunity lies in co-creating unique programmes that provide students with
real-world immersion, where industry experts teach, co-develop case studies, and expose students to actual tools, challenges, and
business scenarios. This blend of academic fundamentals with practical, current insights is incredibly valuable.
Your own perspective as a student aligns with this — when someone from industry walks into a classroom and explains what is
happening right now, learning becomes far more relevant and actionable. This partnership will only strengthen with time. As AI
continues to evolve, academia and industry coming together will be essential for shaping a truly future-ready workforce.
With GenAI and LLMs rapidly expanding globally, how do you envision them shaping the future of AI and impacting businesses?
There are countless narratives and perspectives, but one undeniable shift is taking place not only in large enterprises but also at the
individual, everyday user level. We are witnessing a massive democratisation of insights and capabilities. Students like you are already
using tools like ChatGPT, Copilot, or cloud-powered assistants for research, assignments, and productivity. The same transformation is
happening across industries, where AI is becoming accessible all the way to the last mile.
AI will soon function like a personal digital companion, an intelligent assistant that helps you think faster, work smarter, and solve
problems that once seemed too complex or time-consuming. At Tiger Analytics, we’re seeing this shift firsthand, which is why our
philosophy today is centred around “AI for Enterprise.” It’s no longer just about AI for decision-making or impact; it’s about AI for data,
processes, support, sustainment, insights, and ultimately, end-to-end business value creation.
AI is poised to permeate organisations both vertically and horizontally. It will make businesses more efficient, more resilient,
and more competitive. In many ways, adopting AI will become a strategic advantage, and not embracing it will turn into a
strategic disadvantage. The change is already unfolding, and over the next few years, AI will become a fundamental layer in
every business function and job role. The future of work and the future of AI are now deeply intertwined.
In your role, you juggle leadership, CXO engagement, and P&L responsibilities. What keeps you inspired and balanced in such a
fast-moving industry?

There isn’t a single formula; it’s been a humbling journey with constant learning. But today, at the scale at which I operate, two
things inspire me the most.
First, the impact on people. When you see colleagues you’ve mentored stepping into larger roles, performing brilliantly, and
driving change across the organisation, it’s incredibly fulfilling. It motivates me to invest even more time in grooming future
leaders.
Second, the impact on clients. When clients write to you about the tangible transformation they’ve experienced, it’s gratifying to
know that the work has genuinely moved the needle. I’ve learned something important: managing your energy matters more than
managing your time. If you keep your energy aligned through wellbeing, structure, and clarity, time manages itself.
Planning days and weeks in a meaningful way reduces stress. And equally important is prioritisation. Not everything can be
done at once. You need to delegate thoughtfully, neither over-delegate nor under-delegate. That’s key to staying sane while still
performing at a high level.
Could you share a challenge or setback from your career that taught you a lasting lesson?
I’d have to go back to the early days of my career. I was an analyst working onsite for a client. We had just built an MVP for a
new customer acquisition model using SAS at that time. One line of SAS code had an error. As a result, the model ultimately
targeted existing customers for a new customer acquisition campaign.
A customer emailed saying, “Don’t you know I’m already your customer? Why are you trying to acquire me?”

It was a small pilot, so the damage was contained, and we resolved it; however, the incident shook me. It was a game-
changer lesson. It taught me that:

No job is too small.
Every output has a real-world consequence.
Quality checks and processes are non-negotiable.
You are accountable not only for impact, but also for the downside.
This lesson has stayed with me throughout my career.
If there’s one key takeaway for listeners on the skills needed to excel in AI and analytics, what would it be?
I’d actually offer two, because both are equally important. The first is that success in AI today is no longer defined by how you do
something; it’s defined by what problem you choose to solve. With so many technologies and tools now available, you no longer
need to be a hardcore coder to make an impact. This is especially relevant for MBA students or professionals in business-facing
roles, where the real value comes from connecting the dots, truly understanding the business challenge, and shaping the right
solution.
The second takeaway is that you shouldn’t think of AI as something you simply “use.” Instead, partner with it. Treat it like a
teammate. Traditional tools like Excel or PowerPoint never taught you how to use them, but AI can actually teach you to become
better and wiser at what you do. It can elevate your thinking, accelerate your work, and augment your capabilities, but only if you
become comfortable with it.
At the same time, it’s important to remember that the fundamentals never go out of style. Strong business acumen, clear
communication, storytelling ability, collaboration, and the skill to make sense of ambiguity will always matter. AI should enhance
your human potential, not replace it. These are fascinating times, and I genuinely believe the next few years will reshape
everything we know about work, technology, and leadership.