How microsoft/agentic-applications-for-unified-data-foundation-solution-accelerator works
This accelerator provides a conversational interface for asking natural-language questions over governed Fabric data, with agent orchestration, streaming answers, citations, and conversation history. The main flow is user → chat UI → API → agent → data-backed response; history is a separate read/write flow. Scenario-specific provisioning and prompt/agent setup are also product capabilities. The repository supports .NET and Python API implementations; available excerpts do not establish which backend the UI selects at runtime, so those links are marked unverified.
Main entry points: App.tsx Application shell · api.ts API client · Chat.tsx Chat experience · ChatController.cs .NET chat endpoint
Components 23
Web experience
- Application shell
src/App/src/App.tsx - Chat experience
src/App/src/components/Chat/Chat.tsx - History panel
src/App/src/components/ChatHistoryPanel/ChatHistoryPanel.tsx - Citation display
src/App/src/components/CitationPanel/CitationPanel.tsx - Conversation state
src/App/src/store/chatSlice.ts - API client
src/App/src/api/api.ts
API services
- .NET chat endpoint
src/api/dotnet/Controllers/ChatController.cs - .NET history endpoint
src/api/dotnet/Controllers/HistoryController.cs - Python API router
src/api/python/app.py - Python chat service
src/api/python/chat.py - Python history service
src/api/python/history.py - Conversation storage
src/api/dotnet/Interfaces/IConversationRepository.cs - Cosmos repository
src/api/dotnet/Repositories/CosmosConversationRepository.cs - Fabric SQL repository
src/api/dotnet/Repositories/SqlConversationRepository.cs
Agent and data
- Agent framework
src/api/dotnet/Services/AgentFrameworkService.cs - Title generation
src/api/dotnet/Services/TitleGenerationService.cs - Foundry agent
- Fabric data foundation
- Azure AI Search
Scenario setup
- Scenario configuration
infra/scripts/post-provision/02_generate_agent_prompt.py - Fabric item setup
infra/scripts/post-provision/01_create_fabric_items.py - Agent setup
infra/scripts/post-provision/04_create_agent.py
Other components
- Business user
Connections 25
- Business user → Application shell: uses
- Application shell → Chat experience: renders
src/App/src/App.tsx - Application shell → History panel: renders
src/App/src/App.tsx - Application shell → Citation display: renders
src/App/src/App.tsx - Chat experience → Conversation state: updates
src/App/src/components/Chat/Chat.tsx - Chat experience → API client: submits query
- History panel → API client: loads history
- API client → .NET chat endpoint: calls
- API client → Python API router: calls
- Python API router → Python chat service: routes to
src/api/python/app.py - Python API router → Python history service: routes to
src/api/python/app.py - .NET chat endpoint → Agent framework: uses agent
- Agent framework → Foundry agent: creates client
src/api/dotnet/Services/AgentFrameworkService.cs - Foundry agent → Fabric data foundation: queries
- .NET chat endpoint → Azure AI Search: builds citations
src/api/dotnet/Controllers/ChatController.cs - .NET history endpoint → Conversation storage: reads and writes
src/api/dotnet/Controllers/HistoryController.cs - .NET history endpoint → Title generation: generates titles
src/api/dotnet/Controllers/HistoryController.cs - Conversation storage → Cosmos repository: implemented by
src/api/dotnet/Program.cs - Conversation storage → Fabric SQL repository: implemented by
src/api/dotnet/Program.cs - Cosmos repository → Fabric data foundation: stores conversations
- Fabric SQL repository → Fabric data foundation: stores conversations
- Title generation → Foundry agent: generates title
src/api/dotnet/Services/TitleGenerationService.cs - Scenario configuration → Agent setup: configures prompts
- Fabric item setup → Fabric data foundation: creates data items
- Agent setup → Foundry agent: creates agent