TFM is a Telegram bot that helps you manage your finances. The main feature is to convert your voice messages into transactions using OpenAI's Speech-to-Text API.
Why?
Today, almost every bank offers its own mobile application for managing finances , but these apps aren’t always user-friendly, especially when it comes to viewing statistics or analyzing your expenses. I wanted to create a simple, convenient tool for managing finances, accessible directly through a messaging platform. Using AI to convert voice messages into text is a fantastic way to streamline the process - for example, you can simply say, "I spent 10 dollars on coffee," and the bot will understand and process it.
What?
You say something like "I spent 10 dollars on coffee" and the bot will process this command and update your financial records accordingly - deducting 10 dollars from your balance and adding a description (e.g., "coffee") to your transactions. The bot can also provide you with a summary of your account.
The bot can handle multiple users, so you don't need to run your own instance.
PS. Currently, the bot is in the early stages of development, and actually I don't have a lot of time to work on it, but I hope to add more features in the future.
How?
Initially, I decided it would be great to use Google Sheets as a database for storing financial records. This approach allows users to easily analyze and manipulate their data as they see fit. However, this requires the bot to access the Google Sheets API, which is slightly more complex than using a traditional database. As a result, each user needs their own instance of the bot, along with a personal Google Service Account.
Later, I decided to switch to a more traditional database, for allowing multiple users to use the bot without the need for a personal instance.
The bot is built using Python and the aiogram library. It uses OpenAI's Speech-to-Text API to convert voice messages into text. For the database, I chose PostgreSQL.