I believe AI is not about buzzwords or borrowed expertise; it’s about rolling up our sleeves, building from scratch, and delivering results that redefine speed and scale.
Chief Technology Officer, Aditya Birla Capital
Artificial Intelligence is everywhere today in boardroom presentations, in investor calls, and in endless strategy discussions. Yet, I often feel we are drowning in buzzwords while missing the real point: delivering tangible outcomes. At Aditya Birla Capital, my focus has been clear cut through the noise, get our hands dirty, and build solutions that actually work.
I often say this to my teams and to industry peers: there’s nothing called an “AI skill.” You can keep searching for it, but unless you roll up your sleeves and build something, you’re just telling stories.
Demystifying AI “Agent” vs. “Agentic”
One of our earliest challenges was simply understanding what AI could realistically do. We struggled with terms like “agent” and “agentic,” which are often used interchangeably. Through trial and error, we drew a clear line: agentic systems are fully autonomous, while agents are designed for automation tasks. That clarity was liberating. It meant we could assess problems honestly: if a challenge could be solved without AI, then we didn’t need to force-fit AI into it.
From 90 Days to 4 Hours: Our AI-Driven Audit Revolution
The example I’m most proud of is the AI-powered audit platform we built completely in-house. A team of just 10 people, none of them with prior AI expertise worked alongside me on this project. Together, we created a system that shrank compliance and audit timelines from 90 days to 4 hours. Today, near real-time compliance reporting is just a button-click away. That project proved to me and to the organization that AI is not about expertise on paper. It’s about curiosity, persistence, and a willingness to learn by doing.
Moving Beyond Hype: Building the Technology Backbone
When everyone around us was rushing into customer-facing chatbots and flashy pilots, I made a conscious decision to focus first on the technology backbone. Without scale and speed at the foundation, no innovation can survive. For the next 18 months, my team’s energy is dedicated to building strong, scalable technology use cases that will empower our business units to innovate with confidence.
The Orchestrator Mindset
I don’t believe the future lies in building massive language models from scratch. Instead, it’s about orchestration. We’ve created a system that can dynamically decide when to use rule-based automation, retrieval-augmented generation (RAG), small language models, or large language models. The goal is simple: optimize for cost, speed, and efficiency.
Lessons for the Industry
If there’s one message I want to leave the industry with, it is this: AI’s promise is real, but so is the knowledge gap. Teams must be trained on tools, not just theories. We need to think like systems architects, not tool collectors. And above all, we must acknowledge the pace of change. Our old estimation tools are obsolete. What once took 40 days now takes 4 hours and that’s the new normal.