Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/davidmpizarro/QuipuEco-Hackaton/llms.txt

Use this file to discover all available pages before exploring further.

QuipuEco is a Progressive Web App that brings intelligent recycling guidance to Lima, Peru. Point your phone at any household waste item and receive instant material classification, step-by-step preparation instructions, CO₂ impact data, and turn-by-turn directions to the nearest Tambo store or Centro Verde municipal recycling point — all powered by Google Gemini AI.

Quickstart

Get the frontend and backend running locally in under 10 minutes.

Architecture

Understand the full-stack architecture: React, FastAPI, and Gemini.

Image Classification

Learn how AI vision identifies waste materials from photos.

Voice Agent

Explore the conversational voice interface powered by Web Speech API.

Recycling Map

Discover the Mapbox-powered interactive map and routing system.

API Reference

Explore the FastAPI backend endpoints for classification and chat.

How QuipuEco Works

1

Capture or Upload a Photo

Use the camera viewfinder or upload an image of your waste item. QuipuEco accepts JPG, PNG, and WEBP up to 10 MB.
2

AI Classifies the Material

The image is sent to the FastAPI backend, which forwards it to Google Gemini’s vision model. Within seconds you receive the material type, estimated weight, CO₂ savings, and handling instructions.
3

Find the Nearest Collection Point

The app opens a Mapbox map filtered to the correct network — Tambo stores for plastic, paper, and metal; Centro Verde municipal points for glass, organic, electronic, and hazardous waste.
4

Ask the Voice Agent Anything

Open the floating voice agent and speak naturally: “How do I prepare this bottle?” or “Where’s the nearest Tambo?” The agent responds in audio and can navigate the app on your behalf.

Key Features

AI Image Classification

Google Gemini vision classifies 8 waste categories: plastic, paper, glass, organic, metal, electronic, hazardous, and non-recyclable.

Conversational Voice Agent

Hands-free recycling guidance using the Web Speech API for both recognition and text-to-speech synthesis.

Tambo+ Integration

Direct drop-off at 6+ Tambo convenience store locations across Lima Este — no appointment needed.

Impact Dashboard

Track your cumulative CO₂ savings, recycled weight, and eco-rank with gamified progression tiers.
QuipuEco was developed by Equipo QuipuEco for the Hackathon Tecsup 2026. The backend requires a Google AI Studio API key and the frontend requires a Mapbox token — see Configuration for setup details.

Build docs developers (and LLMs) love