Topic: Agentic AI in Energy Operations: Where Traditional Systems Meet Autonomous Intelligence
AI is becoming important across many industries, including energy. It can support areas such as equipment monitoring, maintenance planning, forecasting, incident investigation, field operations, and operational decision support.
Agentic AI goes beyond providing answers. An AI agent can understand a goal, create a plan, select approved tools, and perform multiple actions. This additional autonomy also introduces new security and governance concerns. A valid user identity does not automatically authorize every agent, skill, tool, or downstream action.
This session examines these challenges from an enterprise AI architecture and security perspective. It introduces the concept of an Agent Control Layer for managing identity, context, permissions, runtime authorization, human approvals, observability, and audit.
The session also references the published research “Zero-Trust Gateway Protocol for Multi-Tenant AI Agents” as a potential implementation pattern for applying least privilege, per-request verification, tenant isolation, and skill-level access control to AI-agent interactions.
Using an energy control-room analogy, the session explains a simple idea: a valid badge may allow someone to enter, but it does not authorize every switch. In the same way, an authenticated AI agent should not receive unlimited authority.
"The smarter and more autonomous an AI agent becomes, the more carefully its access and actions must be governed."
Date and Time
Location
Hosts
Registration
-
Add Event to Calendar
- Houston, Texas
- United States
- Starts 25 July 2026 05:00 AM UTC
- Ends 17 September 2026 05:00 AM UTC
- No Admission Charge
Speakers
Venkata Phani Patelkhana
Agentic AI in Energy Operations: Where Traditional Systems Meet Autonomous Intelligence
AI is becoming important across many industries, including energy. It can support areas such as equipment monitoring, maintenance planning, forecasting, incident investigation, field operations, and operational decision support.
Agentic AI goes beyond providing answers. An AI agent can understand a goal, create a plan, select approved tools, and perform multiple actions. This additional autonomy also introduces new security and governance concerns. A valid user identity does not automatically authorize every agent, skill, tool, or downstream action.
This session examines these challenges from an enterprise AI architecture and security perspective. It introduces the concept of an Agent Control Layer for managing identity, context, permissions, runtime authorization, human approvals, observability, and audit.
The session also references the published research “Zero-Trust Gateway Protocol for Multi-Tenant AI Agents” as a potential implementation pattern for applying least privilege, per-request verification, tenant isolation, and skill-level access control to AI-agent interactions.
Using an energy control-room analogy, the session explains a simple idea: a valid badge may allow someone to enter, but it does not authorize every switch. In the same way, an authenticated AI agent should not receive unlimited authority.
"The smarter and more autonomous an AI agent becomes, the more carefully its access and actions must be governed."
Biography:
Venkata Phani Patelkhana is an IEEE Senior Member and Distinguished Engineer at Dell Technologies, with more than 19 years of experience in enterprise API platforms, distributed systems, API security, and agentic AI architectures.
His current work focuses on AI control planes, MCP gateways, AI Security, agent identity, runtime authorization, and helping enterprises transition safely from API-first to agent-first ecosystems. He holds patents related to decentralized API gateway architectures and regularly writes and speaks about practical approaches to securing and governing enterprise API and AI ecosystems.
