Practical overview
When it fits
Use it when you want a .NET starting point for RAG over documents stored in Azure SQL Database or SQL Server 2025, with a web interface, API access and Azure OpenAI orchestration.
Who it is for
- Developers building document-question-answering applications backed by SQL vector search.
- Teams that want a Blazor chat interface and an API for querying indexed content.
Documented uses
- Import PDF, DOCX, TXT and Markdown documents and store their embeddings as SQL vectors.
- Search indexed content semantically and generate answers grounded in retrieved document chunks.
- Expose document import and question-answering workflows through a Minimal API.
- Show source citations and token usage alongside generated answers.
Requirements & limits
The repository is public and specifies the MIT license. Running the application requires the .NET 10 SDK, a supported SQL database and Azure OpenAI resource credentials.
- The supported database choices are Azure SQL Database and SQL Server 2025 or later, using native VECTOR storage.
- The application depends on Azure OpenAI for embeddings and chat completion.
- Tokenizer ModelId configuration is separate from the Azure OpenAI Deployment values used for service calls.
- The web chat and /api/ask-streaming endpoint support streamed answers; citations include page numbers when available.
Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.
Set up the application locally
Configure and start the Blazor app and Minimal API against a supported SQL database and Azure OpenAI.
Before you start
- .NET 10 SDK
- Azure SQL Database or SQL Server 2025
- Azure OpenAI resource and API keys
Steps
Clone the project repository.
Edit SqlDatabaseVectorSearch/appsettings.json with the SQL connection string and OpenAI settings.
Run the project with dotnet run --project SqlDatabaseVectorSearch/SqlDatabaseVectorSearch.csproj.
Open https://localhost:7025, or use the port reported in the console.
Expected outcome
The application starts locally with its Blazor interface and API available.
Guide limitations
- ChatCompletion and Embedding ModelId values must be recognized by Microsoft.ML.Tokenizers; service calls use the configured Deployment values.
- Embedding dimensions must match the chosen model configuration. If changing the VECTOR size, update both ApplicationDbContext and the initial migration.
Compare options for Vector search
These projects document a common task and share a project role. This is a source-based comparison, not an execution benchmark or a guarantee of interchangeability.
| Project | Documented purpose | Platform | Access and requirements |
|---|---|---|---|
| SQL Database Vector Search | Import PDF, DOCX, TXT and Markdown documents and store their embeddings as SQL vectors. | Cloud, Web | The repository is public and specifies the MIT license. Running the application requires the .NET 10 SDK, a supported SQL database and Azure OpenAI resource credentials. |
| Chatbot with PDF for Semantic Search | Prepare document content as embeddings and save it to a supported vector database. | Web | The repository is identified as open source and publicly visible on GitHub. |
| AI LLM Playground | Ask questions about PDF documents through an assistant interface. | Web |
Sources & review
Source record details
- Source availability
- Active
- Sources checked
- 12 October 2026
- Original website title
- marcominerva/SqlDatabaseVectorSearch · Official homepage title · retained in its source language