
Superchats
Project / product name: wast_collection Telegram ChatBot
Link to the project: http://t.me/wast_collectio_bot
Team leader: Etienne Duplessix
Challenge: 6. Sustainable AI
Problem: People often don't know how to correctly sort waste or where to recycle it. At the same time, cities have open data on recycling locations that remains underused. This leads to missed opportunities for improving recycling behavior and sustainability in urban environments.
Solution: We built a Telegram chatbot that uses AI to recognize waste from a photo and tells users which bin to use. It then finds the nearest recycling location using open data, solving the common problem of knowing how to sort waste but not knowing where to take it.
Impact: Our solution makes recycling easier and more accessible for everyone. It increases awareness of proper waste sorting, promotes the use of existing open data, and helps reduce environmental harm by guiding people to the right bin at the right place — all through a simple, everyday tool like Telegram.
Feasibility & financials: Public model: The app can be maintained by developers with support from Prague City Hall or sustainability grants, using open data and minimal infrastructure.
Private model: A Telegram subscription can unlock premium features (e.g. statistics, business use), creating a self-sustaining path with low operating costs.
What is new about your solution?: We use a lightweight AI model and easy integration with open data to create a user-friendly solution. Unlike other tools, our implementation is simple, fast, and accessible via Telegram — no app install needed. Competing products lack this combination of usability, data use, and practical AI.
What you have built at the hackathon - text explanation + code (e.g. GitHub link): We built a working Telegram bot that uses AI to classify rubbish from a photo and guides the user to the nearest suitable recycling bin in Prague using open city data. It integrates Google Cloud Vision, Golemio API, and geolocation via Telegram.
GitHub: https://github.com/etienneduplessix/pragueaidiscovery/
What you had before the hackathon, please mention open source as well: Python Light Docker setup
What comes next and what you wish to achieve: We plan to expand to more cities. We aim to collaborate with municipalities and recycling companies to help more people recycle correctly and easily — turning open data into daily action.