🤖 Business Assistant Chatbot

An intelligent AI-powered business assistant chatbot that can understand documents, answer queries, and assist with day-to-day business tasks. Powered by LLMs, RAG (Retrieval-Augmented Generation), and vector search.

🧱 Tech Stack

🧠 LLM: llama-70b (hosted using groq), Google-gen-ai(embedding-001)
📚 RAG: LangChain
📦 Vector Store: Pinecone
📄 Document Ingestion: PDF, DOCX, CSV support
🖥️ Backend: FastAPI
🧑‍💻 Frontend: Streamlit

📂 Folder Structure

├── client
│   ├── app.py
│   ├── components
│   │   ├── chatui.py
│   │   ├── history_download.py
│   │   └── upload.py
│   ├── config.py
│   ├── requirements.txt
│   └── utils
│       └── api.py
└── server
    ├── logger.py
    ├── main.py
    ├── middleware
    │   └── exception_handlers.py
    ├── modules
    │   ├── llm.py
    │   ├── load_doc.py
    │   ├── query_handlers.py
    │   └── vectordb_load.py
    ├── requirements.txt
    ├── routes
    │   ├── chat.py
    │   └── upload_files.py
    └── upload_dir

⚙️ How to Use

1. Clone the Repo

##Clone the repo
git clone https://github.com/jkmathuriya/RAGchatbot.git
##Change path
cd RAGchatbot

2. Install Requirements

cd server
pip install -r requirements.txt

3. Set Environment Variables

Create a .env file:

GOOGLE_API_KEY="api_key"
GROQ_API_KEY="api_key"
PINECONE_API_KEY="api_key"
PINECONE_INDEX_NAME='index-name'

4. Run the Backend Server

cd server
uvicorn main:app --reload

5. Edit URL in client/config.py & Run client

#Edit client/config.py with API_URL with backend URL
cd client
streamlit run app.py

🧑‍💼 Use Cases

• Internal company knowledge base
• HR policy Q&A bot
• Document search assistant for sales/legal teams
• Personal business assistant for founders/execs