LLM Stack

A curated collection of LLM application prototypes built while exploring RAG, prompt engineering, and conversational AI patterns. Each subfolder is a standalone mini-project you can run independently.

Projects

Folder	Description	Key tech
Q&A Chatbot USing LLM	Streamlit Q&A chatbot over documents	LangChain, OpenAI
chatmultipledocuments	Chat with multiple PDFs	LangChain, embeddings
pdf_query	PDF question-answering notebook	LangChain, Jupyter
Prompt-Engineering-LangChain	Prompt engineering patterns	LangChain, Jupyter
Text summarization	Text summarization pipeline	OpenAI, LangChain
celebrity_search_application	Semantic celebrity image search	Embeddings, Python
Image_Retrieval_System	Image retrieval with vector search	CV + embeddings
Conversational Q&A Chatbot	Multi-turn conversational bot	LLM, Python
Blog Generation	AI blog post generation	LLM prompting
LLM Generic APP	Generic LLM app template	Python, Jupyter

Quick start

Each subproject has its own dependencies. Typical setup:

cd "Q&A Chatbot USing LLM"
cp .env.example .env   # add your API keys locally  -  never commit .env
pip install -r requirements.txt
streamlit run app.py

Security: .env files, virtualenvs, and API keys must stay local. See .env.example for required variables.

Why this repo

Useful as a reference library for common LLM patterns: document Q&A, summarization, prompt engineering, and retrieval - before graduating to production stacks like PydanticAI + MCP.

Related

• Robinhood AI Portfolio Copilot - production agent architecture
• YouTube Video Summarizer - video summarization app