Sumedh Habbu is a Digital Transformation Consultant at Supreme Industries, a LinkedIn Top Voice, and a passionate advocate for ethical, sustainable, and inclusive innovation. With international experience spanning the U.S. and India, he brings a nuanced perspective on implementing emerging technologies like Generative AI and Retrieval-Augmented Generation in traditional sectors. Sumedh emphasizes the critical role of continuous learning, data stewardship, and contextualized adoption to drive meaningful change. Whether fostering internal AI literacy or exploring voice interaction in LLMs, his approach balances technical vision with grounded responsibility, making him a thoughtful leader in the evolving landscape of digital transformation.

The Social Digest: Navigating the Return: From the U.S. Back to India: After spending nearly four years in the U.S. for your master’s and professional experiences, you’ve recently returned to India. How has this transition influenced your perspective on digital transformation and innovation in the Indian context? 

Oh, a great question right off the bat! To begin with, when you return home after spending a prolonged period in another country, the reverse cultural shock hits you immediately and quite hard. You start noticing the differences the second you step out of the aircraft. But you also begin to notice the progress that has been made in your absence and that fills you with hope.

Coming back from the U.S., a global superpower in technology and innovation, you feel an innate urge to make improvements in systems and processes, to introduce new technologies, and help your country vault forward into a leadership position. You see so many fantastic things happening there, and you want to bring those home. You want your country to succeed.

But that’s when you start to encounter friction. As Indians, we have a different style of working. I feel the key to digital transformation here lies in recognizing these differences and learning how to balance them. We have an abundance of talented people, so understanding and adopting innovation isn’t the hard part. The real challenge is building infrastructure and making resources accessible to that skilled human capital. Striking a balance between driving social change and introducing new technology is paramount to meaningful digital transformation in India.

For example, consider electric vehicles (EVs). Until 2023, you’d rarely spot one on the street. But when I landed in Mumbai in February this year, I was surrounded by EVs – both two- and four-wheelers. You could barely spot a pure ICE vehicle. This perfectly encapsulates how transformation – digital, social, economic – unfolds in India: initial reluctance to adopt, followed by a sudden leap to mass usage. We saw this with the internet, mobile phones, and now we are seeing it with AI. From where I stand, I can say with confidence that we are ready to take the next step in digital transformation. 

The Social Digest: Emphasizing Learning Culture in AI Integration: You’ve conducted sessions on AI basics to foster a learning culture within your organization. How do you see the role of continuous learning in successful digital transformation, especially in industries not traditionally associated with rapid tech adoption? 

Well, as they say – practice makes perfect! Going beyond aphorisms, let us first define the landscape. When we talk about industries not typically associated with rapid tech adoption, we often misinterpret their actual setup. I would simplify this as a distinction between IT and non-IT companies. But this does not mean non-IT firms are Luddites. They might not have SDEs working in Rust or Kotlin instead of Java or Python, but they often use state-of-the-art machines, systems, and processes. It may sound rudimentary, but that distinction needs to be made.

With that context: In digital transformation, 50% of the challenge lies in making the change stick. Too often, people revert to older systems simply because they are more comfortable with them, which is only human. But the whole point of digital transformation is to enable exponential growth through thoughtful change. That goal gets lost if new systems are not genuinely adopted. IT companies have an edge here. They are already immersed in a software-heavy environment and used to constant change, so their learning curve is shorter. Non-IT companies, on the other hand, must completely rethink the way they work. And the only real way through is continuous learning. Otherwise, it just ends up being another workshop with someone speaking polished English, showing colourful slides on a projector. Thankfully, I am now observing a marked change in people’s outlook – they are embracing transformation far more eagerly. And hey, if AI undergoes reinforcement learning, why shouldn’t we? 

The Social Digest: In a recent post, you discussed experimenting with voice prompts for AI assistants. What insights have you gained from this exploration, and how do you foresee voice interaction shaping the future of AI usability? 

Multimodal LLMs are deservedly all the rage these days. We have gone from using simple text inputs to feeding in images, code, and even voice. Voice prompts, however, have not yet gained mainstream traction, and there’s a simple reason for that: we don’t yet have fine control over the resulting input. That’s why most of us still prefer typing rather than using voice-to-text (VTT) features on our phones. The software still makes mistakes.

But LLMs handle voice inputs differently. Unlike sending a text message via VTT, where you still review what’s typed, here you are speaking directly to the LLM. There’s no intermediary step, no manual correction or confirmation. That makes the interaction more fluid, but also more prone to misinterpretation if your phrasing isn’t precise. In a way, it forces you to think clearly and speak deliberately, almost like you would in a real-time conversation with another person. And that is what makes it both exciting and slightly unpredictable.

It is fascinating to observe and participate in this evolution. Just recently, I noticed Google’s Gemini getting significantly better at parsing voice than it did a few weeks prior. As LLMs continue to train and evolve, I believe voice interaction will become second nature, like chatting with a friend. 

The Social Digest: You’ve shown interest in data stewardship and ethical data collaborations. How do you approach these considerations in your projects, and what frameworks or principles guide your decisions? 

Right, so, this is where the rubber truly meets the road. Data stewardship is not just a buzzword to toss around in boardrooms. It is the bedrock of any meaningful and sustainable digital transformation. My approach is simple: treat data the way you would treat someone’s personal diary. You wouldn’t leave it lying around, share it with strangers, or use it for something the owner never agreed to, right? The same applies to organizational data. Every piece of data has a story, a purpose, and most importantly, an owner – whether that’s a customer, employee, or business partner.

One brilliant guide for people on this journey is called the Three Pillars framework: Transparency, Accountability, and Purpose Limitation. 1. Transparency means people should know what data you are collecting, how you are using it, and who has access to it – no hidden agendas or fine-print tricks. 2. Accountability ensures there is always a clear chain of responsibility. If something goes wrong, someone must take ownership. 3. Purpose Limitation means data collected for one specific reason should not quietly end up being used for something else, at least not without proper consent.

Here’s the thing – ethical data practices aren’t a constraint on innovation; they’re actually an enabler. When people trust you with their data, they engage more fully with your systems, provide better quality information, and ultimately help you build superior products. It becomes a virtuous cycle. Take Apple’s use of differential privacy, for instance. They are able to gather valuable behavioural insights without compromising individual privacy. That’s ethical innovation at its best: getting your cake and eating it too! The challenge, especially in the Indian context, is that awareness around these ideas is still evolving. But we are at an inflection point. I genuinely believe that data ethics will soon become as important as cybersecurity. Not a side note, but a foundational requirement.

The Social Digest: Reflecting on your observations about companies like Snowflake leading in AI/ML adoption, what lessons can other organizations learn from their approach to integrating advanced technologies?

The first lesson organizations must internalize is this: in today’s world, you cannot afford to stay stuck in your comfort zone when it comes to technology adoption. Especially in India, where even parts of the “organized” sector operate quite unorganized. Yes, learning from the West is useful, but we need to contextualize those lessons locally. And before we start chasing advanced AI/ML capabilities, can our businesses first get a proper website up? A social media presence? We have a lot of catching up to do, and moving slowly will not help. We need to leap, not crawl.

Adopting emerging technology is inherently risky and AI/ML is even more so, given how fast it evolves and how much hype surrounds it. But we must let go of our mass-adopter mindset and move toward being early adopters. That is the second big lesson. We lose out on precious first-mover advantage by hesitating. Once implemented, the learning curve for AI/ML is often surprisingly short and usage usually spikes. But identifying the right use cases is the real hurdle. And here, we face the opposite problem: decision-makers get swept up in hype and want to implement AI by hook or by crook.

The issue with AI is that it seems like a universal key, when in reality it isn’t. You must carefully evaluate whether and how it fits your organization. We have already seen examples where companies rushed to replace human roles with AI (particularly in customer service) only to reverse course later. The technology did not deliver the nuanced support they expected, and they had to rehire those human positions. This wasn’t necessarily a failure of AI, but rather a miscalculation of its readiness and fit. It is a reminder that while AI is powerful, it still needs humans in the loop.

The Social Digest: Balancing Technological Innovation with Sustainability: You’re passionate about tech, sustainability, and ethical innovation. How do you balance the drive for technological advancement with the imperative for sustainable and responsible practices? 

My personal motto is: Planet, People, and Profit. There is absolutely no reason these three cannot coexist. For example, LLMs and other AI/ML models require massive datasets to train. You can either pay exorbitantly for small chunks of curated data, or you can – rather unethically –scrape it. To resolve this, we are seeing the rise of synthetic data, which is data that mimics real-world characteristics without containing real information. It is an elegant, ethical solution to a sticky problem.

Similarly, we need checks and balances for AI/ML usage. India is already stepping up with the DPDP Act of 2023, which seeks to balance individual privacy with legitimate data processing. NITI Aayog launched the National Strategy on Artificial Intelligence in 2018. The EU has GDPR, Australia has its Privacy Act, and the U.S. has various state and federal laws. We are moving in the right direction.

Often, the choice is not between right and wrong – it is between what is right and what is easy. And while there is no such thing as a perfect product, we should not compromise ethics for speed. Then there is the question of sustainability. Data centers consume massive amounts of electricity, water for cooling, and release enormous heat. Companies are responding with old-school and new-school ideas. Microsoft and others are investing in nuclear energy. They have also experimented with undersea data centers, and found coral growing on the server containers! An unexpected, serendipitous win for nature. All of this shows we can operate in an environment that’s ethical, regulated, sustainable, innovative, and still profitable. 

The Social Digest: For professionals aspiring to lead in digital transformation, what key skills and mindsets do you believe are essential for success in this rapidly evolving field?

Digital transformation is a broad-spectrum field with many entry points. Essentially, there are two main types of roles: strategists and implementers. Strategists identify the problem, set the product vision, and pitch the solution. Implementers build, manage, and integrate that solution into existing systems. Both are crucial, but the scope of work differs. If you are aspiring to enter this field, I would recommend gaining 4–5 years of professional experience first. Read constantly. Keep up with market trends. And do not underestimate the value of networking – you have to be comfortable building professional relationships.

But here is something that is often overlooked: mindset. Digital transformation isn’t just about tools and technology – it’s about leading change in environments where inertia is strong. That means dealing with skepticism, navigating politics, and facing resistance with patience and clarity. The best leaders in this space are not just tech-savvy, they are resilient, curious, and empathetic. If you can bring those qualities to the table, you won’t just survive in this field, you’ll thrive. The ultimate goal is to help people and businesses adopt, adapt to, and benefit from meaningful change.

This interview was conducted by Manav Vala from The Social Digest on 30/04/2025. If you have any interview recommendations or have a story that you want to share with our readers, get in touch with our editor Vedant Bhrambhatt, at editor@thesocialdigest.com