LedgerSmart LLM application to analyge general ledger data

LedgerSmart is a RAG app based on OpenAI LLM using Pandas AI to interact and visaulize ledger data. This application is piblished on streamlit for demo purposes. This app shows how using Pandas AI we can leverage Already existing pandas dataframes in our system. We can give prompts to our local data in CHAT-GPT style and the application give you the required result by converting the Natural language to Pandas Code. This application can be modified as per the requirement and business needs.

Uses

• SmartExplorer: User Prompt based data analysis and retrieval

• SmartGraph: User Prompt based data plotting

Installation

1. Clone the repository

git clone https://github.com/kushagra1331/ledgerllm.git

2. Create a Python environment

Python 3.6 or higher using venv or conda. Using venv:

cd ledgerllm
python3 -m venv env
source env/bin/activate

Using conda:

cd ledgerllm
conda create -n venv 
conda activate venv/

3. Install the required dependencies

pip install -r requirements.txt

4. Set up the keys in a .env file

First, create a .env file in the root directory of the project. Inside the file, add your OpenAI API key:

OPENAI_API_KEY="your_api_key_here"

Save the file and close it. In your Python script or Jupyter notebook, load the .env file using the following code:

from dotenv import load_dotenv, find_dotenv
load_dotenv(find_dotenv())

By using the right naming convention for the environment variable, you don't have to manually store the key in a separate variable and pass it to the function. The library or package that requires the API key will automatically recognize the OPENAI_API_KEY environment variable and use its value.

When needed, you can access the OPENAI_API_KEY as an environment variable:

import os
api_key = os.environ['OPENAI_API_KEY']

Now your Python environment is set up, and you can proceed with running the experiments.

Reference

https://docs.pandas-ai.com/en/latest/