In Episode 157 of Stay Sharp in Digital Engineering with Razorleaf, co-hosts Juliann Grant and Jonathan Scott welcome back Prithvi Singh Chauhan, Machine Learning Engineer at Geminus AI, for a closer look at the company, its computational autonomy platform, and the future of AI-driven engineering decisions.
Geminus AI takes a different approach to industrial AI. Rather than replacing the physics-based simulators engineering teams already trust, its platform adds a decision layer that helps turn complex simulations into faster, actionable operational recommendations. By connecting existing models, operational data, and multiple physical assets, Geminus aims to help organizations make informed decisions in real time—with physics-based accuracy and an understanding of uncertainty.
What You'll Learn in This Episode:
- Where Geminus AI fits in digital engineering: How its computational autonomy platform bridges the gap between traditional simulation and real-time operational decision-making.
- Why physics-based AI matters: Why high-consequence systems require more than conventional pattern recognition, particularly when decisions affect safety, production, cost, and sustainability.
- How faster decisions deliver business value: Hear how Geminus helped an energy operator develop pipeline and production diversion recommendations in just a few days after regional conflict damaged part of its network—a process that traditionally could have taken weeks.
- How the platform works: Explore its four core capabilities: reasoning, model building, model calibration and drift detection, and system-of-systems optimization.
- How Geminus builds trust: Learn how operators can validate recommendations against the high-fidelity simulators they already use.
- What it takes to deploy the technology: Discover how the platform works with existing simulators and data, supports cloud and on-premises environments, and accommodates different engineering and operational roles.
- Where computational autonomy is headed: Explore the vision for connecting assets across entire enterprises and enabling faster decisions around production, energy resilience, and sustainability.
From Simulation to Operational Decisions:
Many industrial organizations already have sophisticated simulation tools and physics-based models. The challenge is making those models fast, current, and connected enough to support decisions as operating conditions change.
Geminus AI's computational autonomy platform is designed to address that challenge. It builds models using trusted simulators, monitors their accuracy against live operational data, incorporates uncertainty into decisions, and connects individual assets into larger systems.
Prithvi shares a real-world example involving a damaged energy pipeline network, where the team developed recommendations to redirect production while maintaining operational constraints and customer commitments. The discussion also explores how model drift can reveal potential physical problems in an asset, making AI useful for maintenance as well as optimization.
Meet the Guest
Prithvi Singh Chauhan is a Machine Learning Engineer at Geminus AI, where he combines machine learning, high-fidelity simulation, and optimization to improve efficiency and sustainability across complex energy systems. His experience spans international energy projects and other high-consequence industries. He holds a master's degree in petroleum engineering from Texas A&M University, is a named inventor on four U.S. patents, and was named the Society of Petroleum Engineers' Young Engineer of the Year in 2026.
Continue the Conversation
For more context, listen to Prithvi's previous Stay Sharp in Digital Engineering with Razorleaf, Episode 154 appearance on operational digital twins, where he discusses the technology behind physics-based AI and its role in complex industrial systems.
Subscribe to Stay Sharp in Digital Engineering with Razorleaf, for more conversations about digital engineering, product development, manufacturing, AI, and the technologies connecting them.
Music is considered “royalty-free” and discovered on Story Blocks.
Technical Podcast Support by Jon Keur at Wayfare Recording Co.
© 2026 Razorleaf Corp. All Rights Reserved.
