Vector search doesn't require a dedicated vector database. PostgreSQL with pgvector gives you similarity search right next to your relational data. Here's how to set it up in .NET with Aspire, Dapper, and Ollama.

MilanJovanovic's platform  covers topics related to software development, technology trends, and coding tutorials. Through articles, tutorials, and code examples, MilanJovanovic offers insights into programming languages, development frameworks, and best practices in software engineering. Readers can learn about software design principles, development methodologies, and emerging technologies to enhance their coding skills and build high-quality software applications.

Milan Jovanović

pgvector is a PostgreSQL extension that adds vector storage and similarity search without requiring a dedicated vector database. This walkthrough covers setting up pgvector in a .NET application using .NET Aspire for infrastructure provisioning, Ollama with the qwen3-embedding model for generating embeddings, and Dapper for database access. It includes creating the vector column, enabling the HNSW index for fast approximate nearest-neighbor search, generating and storing embeddings, and querying by cosine distance. The IEmbeddingGenerator abstraction from Microsoft.Extensions.AI keeps the embedding provider swappable.

Getting Started With PgVector in .NET for Simple Vector Search

Provisioning Infrastructure With .NET Aspire