Skip to content

Latest commit

 

History

History
35 lines (27 loc) · 851 Bytes

README.md

File metadata and controls

35 lines (27 loc) · 851 Bytes

🤖 Chat with Databases

🤖 Text to SQL Agentic langgraph chatbot

Streamlit Text to SQL Agentic ChatBot app built with langgraph workflow :

  • Workflow : LangGraph Workflow with text-to-query, sqlite, and memory & session management
  • Inference & LLM : Groq Inference, Model : llama3.2 3b

Interface :

ChatBot App

Steps :

  • Put GROQ_API_KEY in .streamlit/secrets.toml
  • Put your csv file in /data folder
  • Use script /utils/csv_to_sqlite.py to create sqlite databases
  • Run the App :
pip install -r requirements.txt
streamlit run main.py

📊 LangGraph WorkFlow

LangGraph Work-Flow

Workflow

⌨ APP DOCKERFILE :

To build the App from Dockerfile

docker build -t chatbot-app .
docker run -p 8501:8501 \
    -e GROQ_API_KEY=your_groq_api_key \
    chatbot-app