This projects implements a REST-API that interacts with a vector database.

API_VDB is a python-based project built with FastAPI and ChromaDB. It enables the user to upload text documents into a vector database and search for the most similar documents.

• Upload documents [txt, md, pdf]
• Delete Documents
• List all uploaded documents
• Text search that returns the most relevant document

• Python 3.12
• Docker Desktop (for container deployment)
• Microsoft Visual C++ 14.0 (for windows deployment, https://visualstudio.microsoft.com/de/visual-cpp-build-tools/)

git clone https://github.com/Gusse-dev/API_VDB
python -m venv env
env/Scripts/activate
pip install -r ./requirements.txt 
# pip install -r ./requirements_windows.txt if running on windows
fastapi run app/main.py

git clone https://github.com/Gusse-dev/API_VDB
docker build . -t api_vdb   
docker run -d --name api_vdb -p 8000:8000 api_vdb 
# docker start api_vdb
docker stop api_vdb

Info: Using the API inside the container for the first time might take a while, as the language model for the vector database needs to downloaded.

View the Swagger-Documentation by running the app and accessing http://127.0.0.1:8000/docs#
It is also provides an interface to test the functionality of the API.
Authorization can be done by using the top right 'authorize' icon and entering a valid username and password.