backonthelangchain

Welcome to backonthelangchain, a personal sandbox for exploring advanced LangChain concepts, including:

This repo showcases experimental and practical use cases for building smarter LLM applications using the LangChain framework and related tools.

Notebooks

Notebook	Description
`github_lookup.ipynb`	Uses Tavily and OpenAI to search for a GitHub user, retrieve their public profile, and summarize it using an LLM + Pydantic output parser.
`react_news_summary_agent.ipynb`	A ReAct-based LangChain agent that fetches recent news on a topic, summarizes each article, and estimates reading level and time using custom tools.
`ica_rag_pipeline.ipynb`	A LangChain + Pinecone pipeline that ingests ICA content from PDFs and YouTube, then answers user questions using retrieval-augmented generation.
`langgraph_reflection_agent.ipynb`	An iterative tweet revision loop using LangGraph. Simulates alternating tweet generation and critique using ChatOpenAI and a branching message graph.
`reflexion_agent.ipynb`	Implements a Reflexion-style research agent using LangGraph and Tavily. The agent answers a question, reflects on its response, and revises using real-time search.
`agentic_rag_flow.ipynb`	A dynamic RAG pipeline using LangGraph with document grading, hallucination detection, and adaptive routing. Inspired by Self-RAG (Asai et al., 2023) and Adaptive-RAG (Jeong et al., 2024).

Name		Name	Last commit message	Last commit date
Latest commit History 20 Commits
README.md		README.md
agentic_rag_flow.ipynb		agentic_rag_flow.ipynb
github_lookup.ipynb		github_lookup.ipynb
graph.png		graph.png
ica_rag_pipeline.ipynb		ica_rag_pipeline.ipynb
langchain1.ipynb		langchain1.ipynb
langgraph_reflection_agent.ipynb		langgraph_reflection_agent.ipynb
react_news_summary_agent.ipynb		react_news_summary_agent.ipynb
reflexion_agent.ipynb		reflexion_agent.ipynb