From Data to Decisions: The Evolving Role of Data Science & AI in Supply Chain

In the latest episode of The Intelligent Leader Podcast, host Shashank Garg sits down with Prateek Shrivastava, Advanced Analytics Manager at Cummins, to explore the evolving role of AI in supply chain management. With career spanning retail analytics at 84.51° to predictive maintenance in manufacturing at Cummins, Prateek has seen firsthand how data-driven decision-making transforms industries.

This conversation goes beyond buzzwords, diving into the challenges of forecasting demand, embedding AI into business workflows, and navigating disruptions with predictive analytics. Here are some of the key takeaways from the episode:

One of the biggest shifts in AI-driven supply chain management is the move from isolated data science models to fully integrated AI solutions. Many organizations still struggle with models that generate insights but fail to drive real action. Prateek emphasizes that the true power of AI is unlocked when predictive models are embedded into business processes.

Take Cummins’ predictive maintenance solution: Instead of merely flagging potential truck failures, their AI system proactively ensures the necessary parts are available before issues arise. This shift reduces downtime, improves efficiency, and builds trust in AI solutions.

Prateek recalls his time at 84.51° when the COVID-19 pandemic caused an explosion in online grocery pickup orders. Traditional demand forecasting models were rendered useless overnight. The key lesson? External factors—such as weather data, public health metrics, or supply chain disruptions—must be integrated into AI models to create more resilient forecasting systems.

Similarly, at Cummins, vast amounts of telematics data are used to predict when trucks might fail. The challenge isn’t just having the data—it’s identifying the most relevant signals within the noise.

Many companies struggle to scale AI initiatives. Prateek shares a simple yet powerful strategy: start small, demonstrate quick wins, and then expand.

Implementing AI successfully requires two key elements:

1. Explainability : AI models must be transparent and interpretable for business leaders to trust and act on their insights.
2. Actionability : Insights should lead to direct business actions, whether it’s automatically replenishing inventory or adjusting logistics strategies.

By embedding AI into existing workflows and ensuring stakeholders understand its value, businesses can avoid the common pitfall of AI projects that never make it beyond the proof-of-concept stage.

The technology landscape for data science has changed dramatically. Prateek highlights how platforms like Databricks have streamlined model development, deployment, and monitoring. Instead of manually building and tuning models, AutoML automates much of the process, allowing data scientists to focus on solving business problems rather than fine-tuning algorithms.

Looking ahead, Large Language Models (LLMs) and Generative AI are poised to transform supply chain operations even further.  AI-driven systems will not only analyze past data but also recommend optimal strategies in real time, enabling organizations to proactively mitigate risks and optimize logistics.

Supply chains today must contend with more than just operational efficiency—they must also adapt to regulatory changes, tariffs, and sustainability goals. Companies like Cummins are using AI to model various supply chain scenarios, ensuring resilience in the face of shifting trade policies.

Additionally, sustainability is becoming a core focus. AI-driven solutions are now being applied to optimize packaging, reduce waste, and track carbon footprints, helping companies move toward their net-zero goals.

AI in supply chain management is no longer just an advantage—it’s a necessity. The key takeaway from this episode? AI should not exist in a silo—it should be deeply embedded into business operations to drive real, measurable impact.

For organizations looking to scale AI successfully, focusing on explainability, actionability, and business integration is critical. And as AI technologies continue to evolve, the businesses that leverage them effectively will gain a significant competitive edge.

Want to hear the full conversation? Listen to the latest episode of The Intelligent Leader Podcast for more insights on AI, analytics, and the future of supply chain management.