GraphRAG App Project using Neo4j, Langchain, GPT-4o, and Streamlit

In this tutorial, I walk you through the development of Graphy v1, a real-time GraphRAG (Graph Retrieval-Augmented Generation) application. Using LangChain, Neo4j, and OpenAI's GPT-4 and text-ada-002 models, I'll show you how to extract knowledge from documents and enable natural language querying over a graph database.
What You'll Learn:
1. Setting up a modular app where users can input their own credentials.
2. Using LangChain's LLMGraphTransformer to convert documents into graph data.
3. Integrating with Neo4j to store and query graph data.
4. Implementing natural language querying using OpenAI's GPT-4.
5. Enhancing the app's UI with Streamlit, including adding a sidebar, logo, and interactive elements.
By the end of this video, you'll have a functional app that allows users to upload PDF documents, extract their content into a Neo4j graph database, and interact with the data using natural language queries.
Graphy-v1 GitHub: github.com/AIA...
Document Buddy App GitHub: github.com/AIA...
Discord Link: / discord
Document Buddy App Video: • Real time RAG App usin...
If you found this video helpful, please like, comment, and subscribe to my channel for more tutorials like this! Your support helps me create more content to help you in your development journey.
Join this channel to get access to perks:
/ @aianytime
To further support the channel, you can contribute via the following methods:
Bitcoin Address: 32zhmo5T9jvu8gJDGW3LTuKBM1KPMHoCsW
UPI: sonu1000raw@ybl
#graphrag #rag #ai

[@artur50]

great! cannot wait till Ollama v2 version :D

[@aibuild1]

Long time fan, thank you so much for the hard work putting these videos together. Incredible learning as always

[@SajjadHussain-d9d]

Fantastic👍. from Lahore...

[@THEINDIANSAVIOR]

Hi brother this project is really cool and amazing. today i started to build this from scratch i hope i'll develop this . Hoping more projects with langchain langraph and Graphdb [Astradb with LLama3.1]

[@EkShunya]

keep em coming

[@mikew2883]

Good stuff! 👏

[@AIAnytime]

Glad you liked it!

[@DARK-fs3rz]

Can u please make a video on hybrid rag (graph + vector) , it will be really useful

[@sneharoy3566]

Interesting video

[@AIAnytime]

Thanks

[@vamsivamsi2903]

Can u please do the same thing using both local models from ollama
It will be really useful

[@santakmohanty612]

Can you please explain how you placed your vector embeddings for each node. For example for the medicine x did you attach another node with that medicine’s embeddings or in that medicine x node itself you attached the embeddings as it’s properties?

[@AA-rd6nm]

tried several ways no luck.. Failed to connect to Neo4j: Could not connect to Neo4j database. Please ensure that the url is correct

[@liviuspinu11]

Can you please make a video about contextual retrieval rag with n8n or flowise.

[@AIAnytime]

ruclips.net/video/C0Nso1MLxa4/видео.html

[@Aspirant00007]

Hi, I want to know whether it can be functional if our company's neo4j database is at other server, which we access through remote desktop connection specifically jumphost. Because we have so many restrictions working on other server.

[@pfbeast]

please create this features full app "Llm (Ollama, openai & other api llm) + multiple pdf + normal RAG (neo4j) + Graph RAG (neo4j) + hybrid search + Reranker + user interface + web search".
can you make this app ?

[@skmahaboobsubhani8508]

add PDF LINK IN DESCRIPTIONN

[@THEINDIANSAVIOR]

Yeah we need that pdf to understand

[@zanefalcao3230]

please do it using gemini or any other open source LLM

[@ibbobud]

Someone didn’t watch the video

[@zanefalcao3230]

@@ibbobud y bro??

[@stailenknife5453]

im also a united fan 🥲