Hello there! :wave:,
In the recent times I’ve been working on an exciting project called ‘Artha Chakra’. It’s a telegram bot which helps you in tracking your expenses.
How did I build it?
Why a telegram bot?
With numerous finance tracking apps available, the simplicity of tracking expenses by sending a text message, without switching apps, seemed really appealing.
Sketching the Bot’s Functionality
Now that the convenience factor is sorted, I wanted to integrate features such as expense addition, daily reminders, alerts upon reaching monthly spending limits, expense categorization, and displaying the balance.
Setting up ‘Artha Chakra’ on Telegram
Creating a bot on telegram was super easy. I simply created a new bot through BotFather, and like a secret password, I got my token to interact with the Telegram API.
Building ‘Artha Chakra’
For building this bot, I chose Node.js and TypeScript as the framework, relying on the node-telegram-bot-api library for smooth sailing. Commands like “/credit” and “/debit” were coded to handle addition and deduction of expenses respectively. A daily reminder was implemented using a cron job when you run “/enable_reminder”. Here’s peek at that snippet:
cron.schedule("0 21 * * *", async () => {
// Checks if the user has enabled a reminder already
const users = await User.find({ reminderEnabled: true });
for (let user of users) {
// Send a reminder to each user who has enabled the reminder
bot.sendMessage(
user.telegramId,
"Don't forget to add your expenses for today!"
);
}
});Deploying ‘Artha Chakra’
Finding the right hosting service was a hassle with many options available like Vercel, AWS, GCP, Azure and Heroku on the table. Balancing complexity and pricing, I settled with Render.com, whose free tier graciously offered 512MB of space!
Handling Data with MongoDB Atlas
Storing user data securely was of prime importance. I used MongoDB Atlas to host my database on the cloud. The three main collections were: Users, Categories, and Transactions, each playing a key role in storing user information, expense categories, and transaction details respectively.
Challenges I Faced
As I was developing this project, I had a few hiccups:
-
Scheduling a user-specific reminder according to different timezones was a challenge. Initially, I used a cron job to query the User DB every minute, but that caused a speed bump, slowing the bot down. As a workaround, I settled for a 9PM UTC reminder, lightening the load on the User DB.
-
The free tier of MongoDB Atlas limits us to only 10 connections at a time. Steered this bottleneck by implementing a RabbitMQ messaging queue to moderate the incoming connections.
-
Hosting the RabbitMQ server on the cloud and scaling the bot to handle multiple users was another hurdle due to pricing considerations.
So what’s next?
Going forward, I plan to:
- Congratulate a user if they stay within their spending limit and suggest adding the surplus to savings.
- Scale the bot to accommodate more users.
- Display monthly expenses by category percentages.
- (Optional) Add a bit of casualness by incorporating an AI chatbot for user interaction.
Why the name “Artha Chakra” though?
I wanted my app to have an Indian touch. In Sanskrit, “Artha” refers to wealth or money, and “Chakra” means wheel or cycle. The name “Artha Chakra” represents the concept of a financial wheel or cycle of money. Just as a wheel turns in a cycle, Artha Chakra aims to help users navigate their financial journey, make informed decisions, and effectively manage their resources.
So, that was my journey with ‘Artha Chakra’, folks! It was a path of learning, stumbling, overcoming, and growing. If you’ve come across with similar projects or have any questions or suggestions, feel free to DM me @gsavitha_.
Wanna check out the code? here’s the link to the repo: GitHub
I have also documented my journey on my Twitter while building it, do follow me @gsavitha_ for more updates.
Until next time, happy coding and financial navigating!