🔍 LLM Document Analyzer

An AI-powered document analysis web app built with Streamlit and OpenAI GPT-3.5-turbo.
Upload PDFs, Word docs, Excel files, CSVs, and plain text — then ask the LLM to summarize, extract entities, answer questions, compare documents, or run any custom query.

---

✨ Features

Capability	Details
Multi-format upload	PDF, DOCX/DOC, XLSX/XLS, CSV, TXT, MD
Multiple files	Upload and analyze several docs at once
Content preview	See extracted text before analysis
Summarize	3–5 paragraph summary
Extract Entities	People, orgs, dates, locations, money
Q&A	Ask any question; LLM answers from document
Sentiment Analysis	Tone & sentiment breakdown
Compare Documents	Similarities & differences across multiple files
Custom Query	Free-form prompt against the document
In-memory processing	Files are never written to disk

---

🚀 Quick Start (local)

1. Clone & set up environment

git clone https://github.com/YOUR_USERNAME/llm-document-analyzer.git
cd llm-document-analyzer

python -m venv .venv
source .venv/bin/activate          # Windows: .venv\Scripts\activate

pip install -r requirements.txt

2. Set your OpenAI API key

Option A — environment variable (simplest):

export OPENAI_API_KEY="sk-..."     # Windows: set OPENAI_API_KEY=sk-...

Option B — Streamlit secrets file:

mkdir -p .streamlit
echo 'OPENAI_API_KEY = "sk-..."' > .streamlit/secrets.toml

3. Run

streamlit run main.py

Open http://localhost:8501 in your browser.

---

☁️ Deploy to Streamlit Cloud (free, public URL)

1. Fork this repository on GitHub (button top-right of the repo page).
2. Go to share.streamlit.io and sign in with GitHub.
3. Click New app → select your fork → set Main file path to main.py → click Deploy.
4. Once deployed, open Settings → Secrets and add:

   OPENAI_API_KEY = "sk-..."

5. Click Save — the app will reboot with your key loaded. Share the generated *.streamlit.app URL with anyone.

---

🤗 Alternative: Deploy to Hugging Face Spaces

1. Create a new Space at huggingface.co/spaces → choose Streamlit SDK.
2. Push this repository to the Space (via the HF git remote or the web UI).
3. Add your key as a Space Secret (Settings → Variables and secrets):
   • Name: OPENAI_API_KEY Value: sk-...
4. The Space will build automatically. Access it at https://huggingface.co/spaces/YOUR_USERNAME/llm-document-analyzer.

---

🗂️ Repository Structure

llm-document-analyzer/
├── main.py              # Streamlit app (all logic)
├── requirements.txt     # Python dependencies
└── README.md            # This file

---

🔒 Security Notes

• API keys are loaded via Streamlit Secrets or environment variables — never hard-coded.
• Uploaded files are processed entirely in memory (io.BytesIO); nothing is written to disk.
• Content sent to OpenAI is truncated to ~12 000 characters (~3 k tokens) to cap cost and latency.

---

🛠️ Customization Tips

Goal	Where to change
Use GPT-4	Change model="gpt-3.5-turbo" in call_llm()
Raise content limit	Change MAX_CONTENT_CHARS in main.py
Add more file types	Extend extract_content()
Use a different LLM provider	Replace call_llm() with your provider's SDK

---

📦 Dependencies

Library	Purpose
streamlit	Web UI
openai	LLM API client
pdfplumber	PDF text & table extraction
PyPDF2	PDF fallback parser
python-docx	Word document parsing
pandas	Excel / CSV parsing
openpyxl	Excel engine for pandas

---

📝 License

MIT — free to use, modify, and distribute.