- One of the major problems of real time content creators are out of data and data at the correct time.
- Even though if they get means, it will take lot of time when dealing with vector search method, when we are going with traditional RAG system.
- So, with the help of ArangoDB and NetworkX, user will have super experience and precise news about what they want.
- I have included the real time Linkedin content posint using n8n and realtime new fetching and storing the graph database.
- For demo purpose, the news is synthetic news, generated by Ollama.
- You can see arangoHackathon.ipynb file inside
- I have commented everything for the reference.
- I have remove the original dataset , because it is 1.44 GB of size, you can view with splitted and updated folder dataset.
- Make sure you have installed docker
- Run the docker compose file for
docker compose up -d
- Make sure you have installed requirements.txt
pip3 intall -r requirements.txt
- And run the server.py
python3 server.py
- Enter into the fronend folder, run
npm i
andnpm run dev
- Create the .env file and fill OPENAI_KEY