
Energy Central How to turn AI into a power plant with Arushi Sharma Frank
Nov 12, 2025
In this discussion, guest Arushi Sharma Frank, a tech investor and advisor with a rich background in energy markets and data integration, delves into the transformative role of AI in the power grid. She shares insights on how AI data centers can evolve from energy burdens into grid stabilizers through innovative orchestration and design. Arushi emphasizes the need for utilities to embrace flexibility, overcome barriers to scaling grid-responsive solutions, and cultivate a new generation of talent to lead the energy sector into the future.
AI Snips
Chapters
Transcript
Episode notes
Data Centers As Orchestrated Grid Assets
- Emerald AI builds software to make data centers modular and grid-responsive by orchestrating compute down to GPU-rack level.
- Closing the communication 'air gap' between utilities and loads lets data centers act like generators in system operations.
Flexibility Versus Dispatchability
- Flexibility at data centers is different from dispatchability; utilities need firm deliverables, not vague load tweaks.
- Combine station power, generators, batteries and workload shifting into a stack to meet dispatch signals at nodal level.
Standardize Protocols And Reference Designs
- Standardize communication protocols and reference designs so grid integration scales instead of being bespoke.
- Use open standards like OpenADR and IEEE 2030.5 and verticalize around common reference designs to unlock capital.

