SQL数据库向量搜索RAG工具

Prerequisites

• .NET 10 SDK
• Azure SQL Database
• Azure OpenAI resource and API keys

Setup

1. Clone the repository

    git clone https://github.com/marcominerva/SqlDatabaseVectorSearch.git
    

2. Configure the database and OpenAI settings - Edit SqlDatabaseVectorSearch/appsettings.json and set your Azure SQL connection string and OpenAI settings. - Important: The ModelId values for both ChatCompletion and Embedding are used for token counting via Microsoft.ML.Tokenizers. These values must be valid model identifiers supported by the tokenizer library (e.g., gpt-4o, gpt-4, gpt-3.5-turbo, text-embedding-3-small, text-embedding-3-large, text-embedding-ada-002). The ModelId may differ from the actual deployment name you're using in Azure OpenAI. For example, for gpt-4.1 and gpt-5 models set the ModelId to gpt-4o for proper token counting. - If using embedding models with shortening (e.g., text-embedding-3-small or text-embedding-3-large), set the Dimensions property accordingly. For text-embedding-3-large, you must specify a value <= 1998. - If you change the VECTOR size, update both the ApplicationDbContext and the Initial Migration.

1. Run the application

    dotnet run --project SqlDatabaseVectorSearch/SqlDatabaseVectorSearch.csproj
    

5. Access the Web App - Navigate to https://localhost:5001 (or the port shown in the console)

How to Use

• Web App: Use the Blazor interface to upload documents, search, and chat with RAG.
• API: Import documents via POST /api/documents and ask questions via POST /api/ask or POST /api/ask-streaming.

#### Example API Request

POST /api/ask
Content-Type: application/json

{
    "conversationId": "3d0bd178-499d-433a-b2bc-c35e488d9e2c"
    "text": "Why is Mars called the red planet?"
}

Example API Response

{
  "originalQuestion": "why is mars called the red planet?",
  "reformulatedQuestion": "Why is the planet Mars called the red planet?",
  "answer": "Mars is called the Red Planet because its surface has an orange-red color due to being covered in iron(III) oxide dust, also known as rust. This iron oxide gives Mars its distinctive reddish appearance when observed from Earth and is the origin of its well-known nickname",
  "streamState": "End",
  "tokenUsage": {
    "reformulation": {
      "promptTokens": 812,
      "completionTokens": 11,
      "totalTokens": 823
    },
    "embeddingTokenCount": 10,
    "question": {
      "promptTokens": 31708,
      "completionTokens": 227,
      "totalTokens": 31935
    }
  },
  "citations": [
    {
      "documentId": "b1870ad7-4685-42a3-576a-08ddb01159d5",
      "chunkId": "749aba1e-0db5-4033-cfa6-08ddb0115da3",
      "fileName": "Mars.pdf",
      "quote": "surface of Mars is orange-red because it is covered in iron(III) oxide",
      "pageNumber": 1,
      "indexOnPage": 0
    },
    {
      "documentId": "b1870ad7-4685-42a3-576a-08ddb01159d5",
      "chunkId": "215e7197-513f-4fbe-cfa8-08ddb0115da3",
      "fileName": "Mars.pdf",
      "quote": "Martian surface is caused by ferric oxide, or rust",
      "pageNumber": 3,
      "indexOnPage": 0
    }
  ]
}

SQL Database Vector Search Sample

A Blazor Web App and Minimal API for performing RAG (Retrieval Augmented Generation) and vector search using the native VECTOR type in Azure SQL Database and Azure OpenAI.

Screenshots

Web API

Limitations & FAQ

• VECTOR column size: Maximum allowed is 1998. For text-embedding-3-large, set Dimensions <= 1998.
• Supported file types: PDF, DOCX, TXT, MD.
• Known Issues: See Issues