RAG PDF Extractor
Developed a full-stack Retrieval-Augmented Generation (RAG) system to extract structured data from unstructured PDF documents using OpenAI's large language model. The project involved building an intelligent document retrieval pipeline and serving structured, verifiable answers via a cloud-hosted web interface.
Tech Stack & Tools:
- Retrieval & Embedding: Chroma, LangChain
- LLM: OpenAI
- Deployment & DevOps: Docker, Azure Container Registry, Azure Web App
Social Media AI Agent
This project features an AI agent that automates social media content creation by generating platform-specific blog posts (e.g., LinkedIn, Instagram) from YouTube videos. The agent produces engaging, humorous, and informative content tailored to the each platform, helping creators and marketers scale their content effortlessly.
Tech Stack & Tools:
- Programming Language & Runtime: Python, Asyncio
- Agent Frameworks: OpenAI Agent SDK
- LLM: OpenAI GPT-4o and GPT-4o mini
- Data Source: youtube-transcript-api
- Frontend: Streamlit
- Deployment & DevOps: Streamlit Community Cloud, GitHub
AI Chatbot
This project showcases a sophisticated AI chatbot built with OpenAI's latest language models, featuring a modern web interface that enables real-time conversations with adjustable creativity settings and model selection. The application demonstrates advanced prompt engineering, conversation memory management, and dynamic parameter control.
Tech Stack & Tools:
- Programming Language: Python
- LLM: OpenAI GPT-4o, GPT-4o-mini, GPT-4-turbo
- Frontend: Streamlit
- API Integration: OpenAI API (Chat Completions)
- State Management: Streamlit Session State
- Environment Management: Python-dotenv
- Deployment & DevOps: Streamlit Community Cloud, GitHub
- Development Tools: Cursor IDE
