Moran Assaf offers insights into Villa Park’s smart city transformation with Edge RAG on Azure Arc, highlighting advanced AI deployment, resilient edge computing, security integration, and practical adoption strategies in partnership with DataON.

Transforming City Operations: Edge RAG and Azure Arc in Villa Park

Overview

The City of Villa Park, California, in collaboration with DataON and Microsoft, has implemented Edge Retrieval-Augmented Generation (RAG) utilizing Azure Arc and Foundry models. This project enables instant, context-rich, and resilient operational experiences, even regardless of network connectivity.

Edge RAG Features Announced at Ignite 2025

  • LazyGraph RAG: Advanced graph-based retrieval delivering higher-quality and more relevant inferencing.
  • High-Fidelity Parsing: OCR support for documents, tables, and images for enhanced search capabilities.
  • SharePoint Server Integration: Direct querying from Edge RAG over SharePoint data.
  • Multimodal Search: Retrieve and process images, rich outputs, and text data.
  • Upgraded Chat UI: Improved user experience for conversational AI interactions.
  • Fully Disconnected Scenarios: Powered by Azure Local, allowing edge operations during network outages.

Villa Park’s Smart City Blueprint

Challenges faced by Villa Park included complex zoning approvals, CEQA compliance bottlenecks, and ADU permit backlogs. Deploying Edge RAG on Azure Local resulted in:

  • Rapid environmental assessments (days reduced to minutes).
  • Automated end-to-end workflows for compliance and permitting.
  • Conversational AI interfaces offering immediate, cited regulatory answers.
  • Full offline operational support and digital resilience.
  • A replicable model for other municipalities seeking modernization.

Technical Highlights

Deep Search Powered by LazyGraph RAG

Villa Park planners utilize dynamic, multi-document reasoning to address complex queries that span numerous regulatory areas. The system intelligently retrieves and combines information from varied sources in real time, reducing compute costs and latency.

Advanced Document Parsing

  • Ingests free-form text, tables, images, metadata
  • Indexes tables with section context and page references
  • Enables detailed, structured data search (e.g., permit statistics, impact scores)

Enhanced Conversational AI

Staff can switch between data-grounded chat and model-only chat modes for general queries, training, or troubleshooting. This aids operational agility and onboarding.

Agentic Workflows & Multimodal Innovations

  • Autonomous workflow execution at the edge
  • User-driven queries over SharePoint (private preview)
  • Performance boosts: sub-15 second responses; 600-image batch in 36 seconds
  • “Bring Your Own Model” support, including OpenAI GPT-4o and other advanced LLMs

Security, Compliance, and Sustainability

  • Data sovereignty: sensitive information remains on-premises
  • Zero-trust architecture linked to Microsoft’s security stack
  • Compliance-ready design supporting city, state, and industry regulations
  • Energy-efficient edge hardware for a reduced carbon footprint

Customer Feedback

Ray Pascua (Planning Manager, Villa Park): AI-driven workflow automation delivers value to resource-limited agencies. While ongoing evolution is needed, the gains in efficiency for planning are substantial.

Howard Lo (VP Sales & Marketing, DataON): Azure Local and Edge RAG proved practical AI value for public sector operations and offer a blueprint for similar deployments elsewhere.

Resources

Conclusion

Edge RAG deployed on Azure Arc empowers city planners—enabling resilient, innovative, and secure smart operations. The Villa Park & DataON partnership is a case study for scalable, secure municipal AI.

This post appeared first on “Microsoft Tech Community”. Read the entire article here