Microsoft Events delivers an advanced overview of Edge AI development with Microsoft Foundry at Ignite 2025, guiding developers on building and deploying on-device and edge AI apps.

Building and Shipping Edge AI Apps with Microsoft Foundry

Session Overview

This Microsoft Ignite 2025 session explores how Edge AI is evolving and how Microsoft Foundry and Foundry Local simplify building, shipping, and scaling AI applications that process data on-device and at the edge. The presentation highlights new features, demonstrations, deployment workflows, and integration with Azure Arc for real-world model delivery.

Key Topics Covered

  • Edge AI Transformation: Shifting intelligence from cloud to device.
  • Microsoft Foundry Local:
    • Develop client apps for on-device AI.
    • Latest cross-platform support including Android (“PhonePe Collaboration”).
    • Custom model fine-tuning using Foundry Local tools.
    • Audio transcription enablement.
  • Deployment with Azure Arc:
    • Deploying, monitoring, and automating AI model placement on servers and edge devices.
  • DevOps for AI Models:
    • GitOps workflows for model management.
    • OCI Registry usage for model artifacts.
    • ORAS tool integration for pushing and updating OCI artifacts.
  • Versioning and Live Demonstrations:
    • Demonstrating new version releases, updates, and effective deployments running on multi-cluster setups.

Chapters Breakdown

  • Handling disconnection scenarios for AI applications.
  • Audio transcription enablement workflows.
  • Android cross-platform expansion through PhonePe.
  • Fine-tuning custom models with Foundry Local.
  • Recap of cross-platform capabilities.
  • GitOps with OCI Registry for streamlined model updates.
  • Push workflows using the ORAS tool.
  • Multi-cluster deployment demonstrations.
  • Summary and session wrap-up.

Speakers

  • Lior Kamrat
  • Sam Kemp
  • Gautham Krishnamurthy
  • Raji Rajagopalan
  • Microsoft Ignite event details: https://ignite.microsoft.com
  • Model deployment demos and session resources: https://aka.ms/ignite25-plans-BestModelGenAISolution
  • Related sessions: BRK331, BRK329

Technical Takeaways

  • How Microsoft Foundry and Foundry Local empower edge AI development.
  • Practical steps for moving AI workloads from cloud to client devices using Azure Arc.
  • Managing lifecycle, versioning, and updates for AI models at the edge with GitOps and OCI Registries.
  • Hands-on demonstrations of deployment workflows for edge and multi-cluster environments.