In the rush to deploy AI, many companies focus on models and flashy demos. For Rajesh Poojari, the real differentiator is disciplined project management and operational rigor. “The technology only performs as well as the processes and infrastructure supporting it,” he says.
The turning point in his career came when a data-driven feature looked perfect on paper but struggled in production. “The model performed well in tests, but users experienced inconsistency,” he recalls. “It wasn’t the algorithm; it was the data. Events arrived late, definitions varied, and ownership was unclear.” That lesson reshaped his approach: data and infrastructure are core product dependencies, not backend details to be addressed later.
Rajesh has seen this challenge repeat across industries. In one case, an AI-assisted predictive feature launched successfully in demos but degraded over time as real-world data patterns shifted. Without monitoring and retraining workflows, the team learned of the issues only after user confidence was affected. “AI products are not build-once-and-ship,” he emphasizes. “Operational support must be planned from the start.”
When evaluating new features, Poojari asks practical questions: Where does the data come from? How reliable is it today? Who owns its quality? How often does it change? Features that influence user decisions require high data readiness; exploratory or assistive tools allow more flexibility. This method keeps teams aligned and prevents overcommitment before foundations are secure.
Collaboration is central to his approach. He involves the data and ML teams early, discussing constraints, dependencies, and expected behavior rather than focusing solely on solutions. “By creating shared language around tradeoffs and failure scenarios, we turn collaboration into partnership instead of negotiation,” he notes.
He also advocates for essential data literacy among product managers. PMs don’t need to be engineers, but understanding how data flows, what affects quality, and what operational requirements AI features introduce enables better decision-making. Comfort with concepts like latency, accuracy, and reliability ensures informed leadership without overstepping technical roles.
Looking ahead, Poojari sees AI infrastructure reshaping the PM role itself. Roadmaps will depend more on data availability, operational readiness, and feedback loops than on isolated feature delivery. PMs who understand these foundations will influence strategy; those who treat them as a black box risk falling behind. “The role is evolving from coordination to informed leadership at the intersection of product, data, and technology,” he says.
For Rajesh, the lesson is clear: success in AI doesn’t come from flashy models alone. It comes from disciplined execution, strong project management, and treating data and infrastructure as essential parts of the product.
