AI TECHNOLOGY ARCHITECTURE

Ted Tschopp — Southern California Edison
Q1 2025

Vision & Guiding Principles

  • AI that is Safe, Responsible, and Aligned with Human Values
  • Transparent, Explainable, and Accountable Systems
  • Sustainable AI Operations

Our North Star

"We design AI that earns trust — not just compliance."

Implementation Timeline

timeline title Responsible AI Roadmap 2025 : Governance Framework & AI Gateway 2026 : Model Hubs & Workforce Enablement 2027 : Autonomous Optimization & Auditable AI

Enterprise Integration Layers

flowchart LR A[Responsible Data Fabric] --> B[AI Gateway] B --> C[Model Hubs] C --> D[Applications & Agents] D --> E[Business-Led Innovation]

Governance Model

  • Policy & Ethics Council
  • AI Risk Registry Integration
  • Audit & Compliance Automation

AI Risk Registry Lifecycle

sequenceDiagram participant BU as Business Unit participant Risk as Risk Team participant Arch as AI Architecture participant Legal as Legal/Compliance BU->>Risk: Identify Use Case Risk Risk->>Arch: Assess Technical Impact Arch->>Legal: Verify Compliance Legal-->>BU: Approve for Deployment

AI4IT in Action

Integrating AI governance, architecture, and delivery pipelines.

Metrics & Outcomes

  • ≥95% AI Use Cases Risk-Registered
  • 100% Model Documentation Compliance
  • +30% Developer Productivity through CoPilots

Example Code


from ai_gateway import ModelHub

def route_request(request):
    if request.type == "sensitive":
        return ModelHub("governed").process(request)
    return ModelHub("general").process(request)

route_request("customer_insight")

Speaker Notes Example

Note: Remind the audience that AI4IT is about operationalizing governance, not slowing innovation.

Closing Thoughts

AI is the new utility.

Our task: ensure it is safe, explainable, and transformative.