A no-frills expense tracker app enabling quick logging of expenses in natural language, using LLMs for parsing and categorization.

Gullak is an expense tracker designed to make expense management as intuitive and effortless as possible. Input your expenses in natural language, and let the integrated AI handle categorization and analysis. With Gullak, you can easily view reports and logs of your historical transactions, providing you with daily and monthly summaries to better understand your spending habits.

I originally developed Gullak for personal use. I had a habit of recording my expenses on Apple Notes — a quick and simple method that just got the job done. However, I needed a tool that could not only store these entries but also parse and visualize them effectively. While Gullak was created out of my own necessity, I decided to open source it to help others who might be seeking similar solutions.

• Natural Language Input: Simply type your expenses as you would in a notebook or notes app.
• Automatic Categorization: Utilizes AI to categorize expenses based on the input text.
• Visual Reports: Generates visual summaries of expenses, providing insights into spending patterns over time.
• Historical Data: Access and review past entries with detailed logs and reports.

To self host Gullak yourself, you can use the provided Docker image. Here are the steps to get it up and running:

docker run --name gullak-app -v gullak-data:/app/expenses.db -p 3333:3333 -d ghcr.io/mr-karan/gullak:latest

This setup will expose Gullak on port 3333, storing the SQLite database at the specified volume location.

Download Shortcut

Gullak supports integration with Apple Shortcuts, enabling you to log expenses quickly and effortlessly using Siri. This workflow is especially useful for hands-free operation, such as logging expenses on the go.

Follow these steps to configure it:

• Download a copy of the shortcut from here.

• Add the shortcut to your library in the Shortcuts app on your iOS device.

• Customize the GULLAK_API endpoint.

  • Open the Shortcuts app and find the newly added shortcut.
  • Tap on the shortcut to open the editing interface.
  • Look for the gullak_api_endpoint variable within the actions list.
  • Edit the variable to point to your specific Gullak self hosted installation.

• Run the shortcut via Siri:

  • After setting up, invoke Siri and use the shortcut by saying a phrase like "Log expense."
  • Narrate the details of your expense, such as "Spent 150 Rs on groceries."

Gullak allows you to customize various settings through a configuration file in .toml format. Please refer to the provided sample config file. Below is a description of each section and their respective settings:

If you prefer to use a different provider like Groq for OpenAI services, you can specify this in the openai section:

[openai]
base_url = "https://api.groq.com/openai/v1"
model = "llama3-70b-8192"
token = ""

Gullak supports configuration overrides using environment variables. This can be especially useful when deploying to different environments or when you need to secure sensitive data like API tokens. Environment variables must be prefixed with GULLAK_.

Here’s an example on how to set the OpenAI token using an environment variable:

export GULLAK_OPENAI_TOKEN=your_openai_api_token_here

To set up Gullak for development on your local machine, follow these steps:

1. Clone the repository:

git clone https://github.com/mr-karan/gullak

2. Build the UI (a Vue3 app):

make build-ui

3. Compile the Go binary:

make build

4. Start the application:

make fresh

Gullak is licensed under the AGPL v3 license.