PostgreSQL & Semantic Kernel: Examples

Introduction

This repository contains examples and sample code for using SemanticKernel with PostgreSQL in general and Azure DB for PostgreSQL in particular.

These samples can be used with any PostgreSQL with the pgvector extension installed. Additionally, it has code to support using Entra Authentication for Azure DB for PostgreSQL. This is not required for other PostgreSQL instances.

Docker setup

Included is a docker-compose.yml file that sets up a PostgreSQL with the pgvector extension installed. This can be used to test the examples in this repository in a local setup. To run, make sure you have Docker installed and run:

docker-compose up

You can now connect to the postgres database on localhost:5432 with the username postgres and password example.

.NET Examples

The dotnet folder contains examples of using Postgres with the .NET SemanticKernel library. See dotnet/README.md for more details.

• Arxiv Query: Demonstrates how to query the arXiv API for papers and process the results.
• Text Search Extensions: Provides extensions for performing text searches on the data.

Python Examples

The python folder contains examples of using Postgres with the Python SemanticKernel library. See python/README.md for more details.

• Vector Store Retrieval Augmented Generation (RAG): Demonstrates how to search the arXiv API for specific papers based on a topic and category, embed the abstracts using Azure OpenAI's embedding service, and store these embeddings in a PostgreSQL vector store for efficient searching. It includes commands for loading data, searching data, and chatting with the model.