A comprehensive framework to create, test, and benchmark Retrieval-Augmented Generation (RAG) pipelines, supporting multiple architectures (e.g., Graph RAG and Agentic RAG), document splitters, embedding models, vectorstores, retrievers, rerankers, and LLM providers, with an interactive Gradio UI and experiment logging.

Langchain essentials

A minimal, modular LangChain project showcasing key components like ChatPromptTemplate, output parsing, retriever setup, and LCEL chaining — ideal for building structured LLM pipelines and RAG systems.

An exploration of advanced RAG retrieval strategies in LangChain, from Multi-Query to Contextual Compression and Reranking for smarter, more relevant results.

Project demonstrating LangChain Retriever capabilities.

Successfully developed a Multi-Domain AI Personal Assistant using LangChain, OpenAI, and Streamlit. The application seamlessly integrates multiple specialized capabilities, including document-based question answering (QA), Python code execution, debugging, explanation and optimization, web search, latest news retrieval, and currency conversion.