updated readme

Author: codinglabsong

.env.example

@@ -0,0 +1 @@
+GOOGLE_API_KEY=YOUR-GOOGLE-API-KEY

README.md

@@ -1,11 +1,59 @@
-# available file formats
-text files, images (with text through OCR)
+# Learn Anything AI Chatbot
 
-# needed for image OCR
-sudo apt update
-sudo apt install -y tesseract-ocr libtesseract-dev
+Learn Anything AI Chatbot lets you query your own documents using a language model. It indexes a folder of files, converts CSV and Excel sheets into a DuckDB database, and performs semantic search with Google Gemini embeddings. An interactive agent built with LangGraph combines retrieval and SQL queries so you can "chat" with your data.
 
+## Features
+- **Multi-format ingestion** – PDFs, Word docs, PowerPoint, Markdown, HTML, and plain text are split into searchable chunks. Images are processed through OCR so their text is also indexed.
+- **Data summarization** – CSV and Excel files are loaded into DuckDB tables. Summary cards for each table are added to the vector index.
+- **Embeddings & retrieval** – Documents are embedded with `GoogleGenerativeAIEmbeddings` and stored in a FAISS index for fast semantic search.
+- **SQL integration** – The agent can issue DuckDB queries over your uploaded spreadsheets. Only `SELECT` and `PRAGMA` statements are allowed for safety.
+- **Persistent conversations** – The ReAct agent from LangGraph saves its history to SQLite so you can resume chats.
 
-DuckDB restrictions for agent to only SELECT/PRAGMA queries. No destructive queries.
 
-Take time to load first time docs
+## Installation
+1. Install system packages needed for OCR (first time only):
+   ```bash
+   sudo apt update
+   sudo apt install -y tesseract-ocr libtesseract-dev
+   ```
+2. Install the Python package and dependencies:
+   ```bash
+   pip install -e .
+   pip install -r requirements-dev.txt  # optional dev tools
+   ```
+
+## Usage
+1. Place the documents you want to search under `data/` directory.
+2. Run the agent. The first run may take a while as it loads and indexes the files:
+   ```bash
+   bash scripts/run_agent.sh --ask "What kinds of files have I provided?" --load_data
+   ```
+3. For later sessions, omit `--load_data` to reuse the existing FAISS index and DuckDB database. If you have added more documents under `data/`, please load them again using `--load_data` for the first run.
+
+## Supported File Types
+- Text documents: PDF, DOCX, PPTX, Markdown, HTML, TXT
+- Images: PNG, JPG, JPEG, TIFF (processed via OCR)
+- Spreadsheets: CSV, XLSX
+
+## Testing
+Run formatting checks and unit tests with:
+```bash
+pre-commit run --all-files
+pytest
+```
+
+## Repository Structure
+- `src/any_chatbot/` – core modules for indexing, tools, and agent
+- `scripts/` – helper script to launch the agent
+- `notebooks/` – example notebooks for experiments
+- `tests/` – unit tests for the indexing and tool utilities
+
+## Requirements
+- Python 3.10+
+- A Google Gemini API key (`GOOGLE_API_KEY` environment variable)
+
+## Contributing
+Contributions are welcome! Feel free to open issues or pull requests.
+
+## License
+This project is licensed under the MIT License.