Crafting user-centred AI innovation
Divey Sharma
Vice President, Data & AI
Incedo Inc.
Crafting user-centred AI innovation
Divey Sharma
Vice President, Data & AI
Incedo Inc.
Artificial intelligence is no longer a futuristic promise; it has become the engine quietly reshaping how industries operate, how decisions are made, and how organisations unlock new forms of value. As enterprises race to keep up with this shift, the real challenge is no longer understanding AI; it is figuring out how to translate its vast potential into measurable, repeatable, and scalable impact. Guiding organisations through this transition is Divey Sharma, Vice President, Data and AI at Incedo Inc. Over nearly two decades in data, analytics, and AI, Divey has seen the field evolve from traditional modelling and rule-driven systems to the rise of GenAI and agentic technologies that can automate entire workflows. His career spans building high-performing AI/ML teams, shaping decision engines for global financial institutions, and now leading Incedo’s AI Centre of Excellence, where he focuses on creating products, platforms, and solutions that make AI not just powerful but practical, usable, adoptable, and relevant to how enterprises really work. In this exclusive conversation with TradeFlock, Divey reflects on his journey through the shifting AI landscape, the execution gaps that still hold organisations back, and the principles he believes will define the next era of enterprise-grade AI.
How has your journey shaped the way you see today’s AI landscape?
When I began my career, AI was not a defined function, and the work lived under the broader label of Analytics. We spent long cycles building models for narrow problems, only to see many of them sit unused because organisations were simply not ready for data-driven decision-making. Watching the field evolve from those early days to where we are now has been a front-row view of how quickly AI can transform itself and everything around it. Technology moved steadily from rule-based systems to machine learning, then to deep learning, and now to GenAI and agentic systems capable of reasoning and automating entire workflows. Through all this progress, the real breakthrough has been the shift to making AI work inside real business environments, where value emerges only when solutions are easy to deploy, natural for teams to adopt, and tightly linked to measurable outcomes. Living through these shifts has strengthened one belief that has stayed constant for me: AI will define the next decade for organisations that can turn innovation into practical, scalable impact.
What does it truly take to operationalise AI at scale inside an organisation?
Many organisations today find themselves stuck in what I often refer to as the Pilot-to-Production Chasm. They experiment enthusiastically, run impressive proofs of concept, and celebrate early wins, yet only a small fraction ever makes it into daily operations. That is why ROI often feels muted and progress is uneven. As I’ve said before, “There’s a lot of AI experimentation happening, but very little actually makes it into day-to-day use.” Closing this gap always begins with a clear understanding of the problem being solved and an equally clear value path that leaders recognise and champion. From there, organisations need a production-grade AI backbone that brings together domain context, enterprise data, robust pipelines, scalable engineering, and dependable governance. When this foundation is strong, adoption becomes far more natural because AI blends smoothly into existing workflows through simple interfaces and clean integrations. Organisations that bring these elements together are the ones that consistently convert pilots into lasting, enterprise-wide impact.
What is one quiet win at Incedo that made you feel a genuine difference was made?
A meaningful milestone over the last few years has been shaping Incedo’s GenAI and Agentic AI vision with a clear focus on building solutions that scale and deliver real impact. That direction led to the creation of Incedo BrainSpark, an Agentic AI platform that provides clients with a unified, standardised way to build intelligent agents and agentic workflows. By bringing deep domain context into LLMs and other AI models, the platform makes AI far more relevant and effective for specific business needs. Seeing mid-tier clients adopt BrainSpark and experience stronger productivity, improved accuracy, and faster outcomes has been especially rewarding. The recognition the platform has received across multiple industry forums has only reinforced its value. The real sense of accomplishment comes from helping shape a capability that has proven genuinely meaningful for both clients and the organisation.
"India needs its own AI vision, not a copy of the US or China."
India has bold AI ambitions. What should the country focus on, and what could the path look like?
India needs an AI vision built on its own realities rather than following the footsteps of other nations. The priority is to build AI that deeply understands India — solutions grounded in our languages, datasets, public systems, and the complexities of a country this diverse — so they can work meaningfully at scale. The second priority is building AI from India for the world by harnessing our talent, domain depth, and cost advantage to create globally competitive platforms and vertical LLMs. Progress has begun, but delivering on this ambition requires a unified push to build the India AI stack, expand talent development, and strengthen collaboration between government and industry. With decisive action, India has a genuine opportunity to build AI that transforms the nation and showcases its innovation on the global stage.
Featured Magazine -
All Magazines -
