📚 Intelligence Assistant (DO33 Bot)

An intelligent assistant that allows users to query PDF and Excel documents using natural language. Built using React + Vite + TypeScript on the frontend and Flask + LangChain + OpenAI on the backend. The system supports contextual retrieval, optional reasoning, and web search.

---

🚀 Features

• ✅ Natural language queries on PDFs and Excel files
• 🔍 Internal document retrieval using vector search (FAISS)
• 🧠 Optional LLM-based reasoning
• 🌐 Optional real-time web search via OpenAI
• 📎 Source file traceability and download support
• 🧾 Feedback system for likes/dislikes with logging
• 🎙️ Voice-to-text input (speech recognition)
• 📁 Modular ingestion: Add new file types easily

---

📁 Folder Structure

DO33\_Final/
├── backend/                     # All backend code
│   ├── app.py                   # Flask API server
│   ├── config.py                # Global config and flags
│   ├── ingestion/               # Data ingestion logic
│   │   ├── ingest\_pdf.py
│   │   ├── ingest\_excel.py
│   ├── vectorstore/             # Stored FAISS indexes
│   ├── retrieval/
│   ├── reasoning/
│   ├── websearch/
│   ├── llm\_response/
│   ├── utils/
│   └── data/                    # Embedded document metadata (pkl, json)
├── Data/                        # Raw files
│   ├── PDF/
│   └── EXCEL/
├── frontend/                    # React frontend
│   ├── src/
│   ├── public/
│   ├── package.json
│   ├── vite.config.ts
│   └── tailwind.config.ts
├── .env
├── .gitignore
└── requirements.txt

---

⚙️ Setup Instructions

1. 🧠 Backend Setup

a. Create a Python virtual environment

python -m venv do33_env
source do33_env/bin/activate  # or do33_env\Scripts\activate on Windows

b. Install dependencies

pip install -r requirements.txt

c. Add OpenAI API Key

Create a .env file in the backend root:

OPENAI_API_KEY=your_openai_api_key

d. Run ingestion scripts

python data_ingestion/generate_embeddings.py

e. Start the Flask server

python backend\app.py

Server will run at: http://localhost:5001

---

2. 💻 Frontend Setup

cd frontend
npm install
npm run dev

Frontend will run at: http://localhost:8080

---

🧪 Example Usage

• Ask: "What is the problem with the XYZ component and how was it solved?"
• Enable reasoning or web search (optional)
• Receive a structured answer with source file buttons
• Click source buttons to open documents
• Leave feedback via 👍 / 👎

---

🛠 Config Options

Inside backend/config.py:

ENABLE_PDF = True
ENABLE_EXCEL = True

---

📦 Dependencies

• Frontend: React, Vite, TypeScript, Tailwind CSS, Shadcn-UI
• Backend: Flask, OpenAI, LangChain, FAISS, Pandas
• Embeddings: BAAI/bge-base-en-v1.5 via HuggingFace
• LLM: GPT-4o or GPT-3.5-turbo (OpenAI)

---

🧠 Future Improvements

• Role-based access and authentication
• Add support for CSV/Docx ingestion
• Summary and analytics dashboard
• Row-level preview for Excel rows used in answers