Aranya Kundu

As the energy sector embraces digital transformation, predictive maintenance powered by machine learning is becoming essential for asset reliability and operational efficiency. At Entergy, Aranya Kundu focuses on developing in-house AI models that help anticipate failures—particularly in transformers—before they happen. This presentation will walk through their journey of building these models, the challenges faced with data quality and security, and how we aligned technical innovation with real-world operational needs.

• How we define clear objectives and build internal models tailored to specific assets like transformers, focusing on minimizing false negatives to prevent unexpected failures.
• Lessons learned from working with messy, high-dimensional operational data and the importance of data engineering in predictive analytics.
• Navigating internal concerns around data sharing, especially in sensitive environments like nuclear facilities, and building secure, compliant AI solutions.
• How collaboration between data scientists and field experts—and early wins—helped overcome resistance and drive adoption of AI tools across the organization.