Ask Research PDF Web-App

• Tech Stack: Python, Langchain, Streamlit
• Website URL: Link
• Github URL: Project Link

About the Application: This Streamlit-powered web app lets users upload PDFs and naturally inquire about its contents. It's particularly handy for researchers: besides querying, users can get summaries of their papers that are generated from carefully designed prompts. This ensures accuracy and relevance.

How It Operates: The app first divides the PDF into smaller, digestible sections. Using OpenAI embeddings, it turns these sections into vector representations. When a query is made, it identifies and selects the sections most relevant to the question and uses the LLM to craft a comprehensive answer.

Final thoughts: It's my initial venture into the domain of LLMs. I'm profoundly intrigued by AI's capabilities and look forward to creating more substantial projects in the future. Additionally, this marks my first successful web deployment.