text-sql-chatbot

Redis Database offers unique capability to keep your data fresh while serving through LLM chatbot

Features in this demo

Text to SQL generation based on the database

Delivering results through semantic caching

Data expiration after 15 seconds to keep cache fresh

Project Setup

Spin up a Redis instance enabled with RedisStack!

The easiest way to is to use a docker image using the below command

docker run -d -p 6379:6379 -p 8001:8001 redis/redis-stack:latest

If you do not want to use a docker image, you can sign up for a free Redis Cloud subscription here.

Set up the project

Download the repository

git clone https://github.com/mar1boroman/text-sql-chatbot.git && cd text-sql-chatbot

Prepare and activate the virtual environment

python3 -m venv venv && source venv/bin/activate

Install necessary libraries and dependencies

pip install -r requirements.txt

Using the project

Update Config

Make sure you update the env file. You need a open ai api key & a langsmith api key to update the config.

mv env_backup .env

vi .env

Load Data

Do not run if not required, below is just to show the workings

python  data/load.py

This creates the cars_database.db (sqllite) database.

If you run this directly, you will get an error since the cars_database.db is already present and there will be unique key constraint violation

Run the app

streamlit run web/app.py