Agentic RAG with LangGraph: AI Blog Search

Overview

AI Blog Search is an Agentic RAG application designed to enhance information retrieval from AI-related blog posts. This system leverages LangChain, LangGraph, and Google's Gemini model to fetch, process, and analyze blog content, providing users with accurate and contextually relevant answers.

LangGraph Workflow

Demo

AI-Blog-Search-Demo-Update.mp4

Features

• Document Retrieval: Uses Qdrant as a vector database to store and retrieve blog content based on embeddings.
• Agentic Query Processing: Uses an AI-powered agent to determine whether a query should be rewritten, answered, or require more retrieval.
• Relevance Assessment: Implements an automated relevance grading system using Google's Gemini model.
• Query Refinement: Enhances poorly structured queries for better retrieval results.
• Streamlit UI: Provides a user-friendly interface for entering blog URLs, queries and retrieving insightful responses.
• Graph-Based Workflow: Implements a structured state graph using LangGraph for efficient decision-making.

Technologies Used

• Programming Language: Python 3.10+
• Framework: LangChain and LangGraph
• Database: Qdrant
• Models:
  • Embeddings: Google Gemini API (embedding-001)
  • Chat: Google Gemini API (gemini-2.0-flash)
• Blogs Loader: Langchain WebBaseLoader
• Document Splitter: RecursiveCharacterTextSplitter
• User Interface (UI): Streamlit

Requirements

1. Install Dependencies:

   pip install -r requirements.txt

2. Run the Application:

   streamlit run app.py

3. Use the Application:

   • Paste your Google API Key in the sidebar.
   • Paste the blog link.
   • Enter your query about the blog post.

📫 Connect With Me