Inspiration
As college students we were tired of keeping track of all the messages on various groups and channels, and more often than not the important stuff gets lost amidst spam. So we wanted a solution that could just tell us what was important and what we needed to know.
What it does
The bot serves two primary features tagging and summarising:
Tagging
It tags all your messages on groups and channels on related topics like "Event, Assignment, Meeting, Computer Science" etc. and then you can select a particular tag and it will display all the messages assigned that tag across all groups and channels.
Summarising
It reads all your messages on the groups and channels and then summarises the conversation for you highlighting only the important stuff you would want to know.
How I built it
The bot was built using the Telegram API and the Python Telegram library, which provided the core framework for creating commands, handling user interactions, and managing inline buttons.
MongoDB was used to store relevant data such as chat messages, user preferences, and tags, allowing for efficient querying and retrieval of information.
To enhance the bot’s functionality, the Gemini API was integrated for summarizing messages and generating topic tags. This enabled users to quickly extract meaningful insights and organize their chats more effectively.
These tools worked together seamlessly, combining real-time interaction through Telegram, reliable data storage with MongoDB, and intelligent processing via the Gemini API.
Challenges I ran into
The only challenge we ran into was that we could not make the bot work for private conversations i.e dms because the telegram api does not allow bots to access private conversations.
Accomplishments that I'm proud of
The one thing that I am extremely enthused about is that we solved a common pain point for many people and I can see many people adopt our solution. I am proud of building a practical product in 24 hours.
What I learned
This was the first bot any of us built, and the learning journey was truly rewarding. We tackled challenges such as managing complex callback structures, seamlessly integrating MongoDB for data storage, and leveraging the Gemini API for advanced functionalities like summarization and tagging—all while working within a tight time frame. We learnt the importance of not giving up when code can sometimes be funky and give unexpected errors (there were a lot of those), or when we couldn't think of implementation of ideas straight away.
Additionally, we explored and experimented with various frameworks to create an engaging and functional landing page, further enhancing the project’s scope and presentation.
What's next for TeleTubby
We will next build a layer of real time Q and A with the bot so you can ask the bot questions about the conversations on the groups and channel and it can give you the answers you are looking for. (Making it your TeleGPT)
Log in or sign up for Devpost to join the conversation.