• Security Defaults
  • Azure AD authentication
  • Secure secrets management
• Resilience & Scalability
  • Automatic retry policies
  • Circuit breakers for failure protection

• Custom Commands
  • Workflow automation
  • Database migrations, data seeding
  • Development environment reset
• Launch Profiles
  • Multiple environments (local, cloud, test)
  • Environment variable configuration

• Real-time Monitoring
• Interactive Debugging
• Resource Management
• Secret Management
• Copilot AI Debugging
• GenAI Visualizer

Hosting Integrations

• Provision and manage resources
• Automate infrastructure setup
• Cloud-native functionality
• Integrated logging and monitoring
• Authentication and authorization
• Configuration management

Client Integrations

• Configure service connections
• Consume hosted services
• Service discovery
• Service defaults
• Integrated logging and monitoring

Cloud-Agnostic

• Databases: PostgreSQL, MySQL, MongoDB, SQL Server
• Messaging: Kafka, RabbitMQ, NATS
• AI & Observability: Ollama, Semantic Kernel, OpenTelemetry

Cloud-Specific

• Azure Services: OpenAI, Cosmos DB, SQL, Redis, Key Vault, App Service, ACR, App Config
• AWS Support: Via Hosting.AWS package

Multi-Language Support

• JavaScript: Bun, Deno
• Systems Programming: Go, Rust
• Enterprise: Java/Spring
• Extensions: Node.js, SQL, MongoDB, Redis

Additional Services

• AI & Search: Ollama, Meilisearch
• Data: SQLite, Data API Builder
• Web: Azure Static Web Apps emulator
• Customization: Advanced configurations

// Local development with Ollama
builder.AddOllama("ollama")
  .WithModel("llama3");

// Add client to use the model
builder.AddOllamaApiClient()
  .AddChatClient();

public class ExampleService(IChatClient chatClient)
{
    // Use chat client...
}

• Zero Cost Development: No API charges for local models
• Privacy & Security: All data stays on your machine
• Offline Capability: Work without internet connectivity
• Fast Iteration: No network latency for testing
• Model Experimentation: Try different open-source models

// Single line change for production
builder.AddAzureOpenAI("ai");

// Same client code works unchanged
builder.AddAzureOpenAIClient()
  .AddChatClient();

public class ExampleService(IChatClient chatClient)
{
    // Use chat client...
}

• Latest Models: Access to GPT-4, Claude, and cutting-edge AI
• Global Availability: 99.9% SLA with worldwide deployment
• Managed Infrastructure: No server maintenance required
• Advanced Features: Function calling, vision, audio processing
• Compliance Ready: SOC 2, HIPAA, and enterprise certifications

Previous Landscape

• Semantic Kernel: High-level orchestration framework
• AutoGen: Multi-agent conversation framework
• Fragmented Ecosystem: Different APIs, patterns, abstractions

The Challenge

• Inconsistent APIs: Each framework had unique approaches
• Integration Complexity: Difficult to combine tools
• Provider Lock-in: Hard to switch AI providers
• Learning Curve: Multiple frameworks to master

Unified Multi-Agent Platform

• Consolidation: Replaces Semantic Kernel & AutoGen
• Built on Microsoft.Extensions.AI: Leverages unified abstractions
• Multi-Agent Orchestration: Coordinate multiple AI agents
• Event-Driven Architecture: Flexible agent communication

Key Features

• Agent Types: Task-based, conversational, autonomous
• Memory Systems: Short-term and long-term context
• Tool Integration: Function calling and external tools
• Provider Agnostic: Works with any AI backend

Learn more: Microsoft Agent Framework

Targets

• Azure App Service
• Azure Container Apps
• Kubernetes

Azure Developer CLI

azd up         # Deploy everything
azd deploy     # App only
azd provision  # Infrastructure only
azd init       # Initialize

Features

• Auto-detects app structure
• Environment variable mapping
• Native Aspire support
• Credential providers
• Key Vault integration
• Secure access

CLI & Tooling

• aspire update - Auto-update packages
• Single-file AppHost (experimental)
• SSH Remote port forwarding in VS Code
• Enhanced aspire exec command

Dashboard Enhancements

• GenAI Visualizer - Explore AI interactions
• Multi-resource console logs view
• Custom resource icons
• Enhanced trace filtering

New Integrations

• OpenAI hosting integration
• GitHub Models & Azure AI Foundry catalogs
• Dev Tunnels hosting support
• Azure Kusto and App Config emulators

Deployment

• Azure Container App Jobs
• Built-in Azure deployment via aspire deploy
• Deployment image tag callbacks

Multiple Environments

var k8s = builder.AddKubernetesEnvironment("k8s");
var compose = builder.AddDockerComposeEnvironment("docker");

builder.AddProject<Projects.Api>("api")
    .WithComputeEnvironment(compose);

builder.AddProject<Projects.Frontend>("frontend")
    .WithComputeEnvironment(k8s);

Aspire CLI

# Install
curl -sSL https://aspire.dev/install.sh | bash
dotnet tool install -g Aspire.Cli

# Commands
aspire new | run | add | update
aspire config | publish | deploy
aspire exec

GitHub Models

var chat = builder.AddGitHubModel("chat",
    "openai/gpt-4o-mini");

builder.AddProject<Projects.ChatService>("chatservice")
    .WithReference(chat);

Azure AI Foundry (Cloud)

var foundry = builder.AddAzureAIFoundry("foundry");

var chat = foundry.AddDeployment("chat",
    "Phi-4", "1", "Microsoft");

builder.AddProject<Projects.Web>("web")
    .WithReference(chat)
    .WaitFor(chat);

Azure AI Foundry (Local)

var foundry = builder.AddAzureAIFoundry("foundry")
    .RunAsFoundryLocal();

var chat = foundry.AddDeployment("chat",
    "phi-3.5-mini", "1", "Microsoft");

builder.AddProject<Projects.Web>("web")
    .WithReference(chat)
    .WaitFor(chat);

Benefits: Privacy, performance, offline operation, cost reduction

// Parameters without defaults trigger dashboard prompts
var apiKey = builder.AddParameter("api-key", secret: true);
var dbUrl = builder.AddParameter("database-url");

var api = builder.AddProject<Projects.Api>("api")
    .WithEnvironment("API_KEY", apiKey)
    .WithEnvironment("DATABASE_URL", dbUrl);

Features

• Automatic prompting for missing parameters
• Rich form inputs in dashboard
• Secret masking for sensitive data
• Validation support with custom rules
• Save to user secrets for persistence

Input Types: Text, Password, Choice, Boolean, Number

Links

• .NET Aspire
• What's new in .NET Aspire 9.5
• Microsoft.Extensions.AI
• Microsoft Agent Framework
• Aspirify
• Aspire Samples
• eShopLite

Follow Chris Ayers

BlueSky: @chris-ayers.com
LinkedIn: - chris-l-ayers
Blog: https://chris-ayers.com/
GitHub: Codebytes
Mastodon: @Chrisayers@hachyderm.io
Twitter: @Chris_L_Ayers