Amazon Annual Report Analyzer
🔬 Overview
This project is an interactive web app that surfaces key themes and strategic focus areas in Amazon’s annual reports. By combining traditional text mining with retrieval-augmented generation (RAG) and LLMs, the app helps users analyze how Amazon narrates its evolving role in society, from retail to cloud computing, logistics, entertainment, and beyond.
The app is particularly useful for students, researchers, and consumers who want to explore how corporate strategy and identity shift through language over time.
🛠 Features
✅ Upload any Amazon annual report PDF
✅ Visualize the top 10 most frequent words by year
✅ Generate guided summaries using top keywords
✅ Ask custom questions using RAG-LLM responses
✅ Experimental: Forecast Amazon’s next moves based on report language
🏗️ Tools & Technologies
- Python + Streamlit (for an interactive web app)
- Pandas + Plotly (data manipulation & visualization)
- LangChain + FAISS (retrieval-augmented generation)
- OpenAI GPT-3.5 (guided summarization and Q&A)
📌 What I Learned
- How to combine keyword-based data mining with LLMs for strategic document analysis
- Designing apps for interpretability, not just answers
🚀 Future Improvements
- Add side-by-side year comparison views
- Integrate sentiment/tone shift visualizations