Terraquest_AI

Ritu Sheelvant — Sun, 11 Oct 2026 18:36:35 +0000

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 1: Touch Grass

What I Built

TerraQuest AI 🌿 — Your world is the game. Go explore it.

TerraQuest AI is a gamified nature exploration and environmental action platform that turns everyday surroundings into real-world adventures.

It features two types of quests:

🌸 NatureQuest: Photograph specific flowers, compare leaf patterns, and capture creatures performing natural behaviors.
♻️ EcoQuest: Complete environmental missions involving responsible litter collection, recycling, waste segregation, and composting.

Users select their surroundings and the flora and fauna they can find. The platform is designed to generate relevant challenges based on their environment, available time, and preferred difficulty.

Complete missions, submit photographic evidence, earn points, unlock levels, and track your progress.

The idea is simple: make the screen the starting point of an adventure, not the destination.

Demo

Coming soon.

Code

TerraQuest AI on GitHub

How I Built It

TerraQuest AI is being developed using Python and Streamlit, with lightweight local storage for mission progress and evidence.

The planned AI integration uses an open-weight language model to personalize outdoor missions according to the user's environment, available flora and fauna, time constraints, and difficulty preferences.

The platform combines personalized quest generation, predefined fallback missions, photo evidence submissions, and a gamification system featuring points, levels, and achievements.

The application is designed to remain usable when an AI provider is unavailable.

Why Does Open Innovation Matter?

Open-weight AI offers an opportunity to make outdoor experiences more adaptable without making a paid, closed-model API a mandatory dependency.

Adaptability: Mission-generation prompts and rules can be modified to suit different surroundings and available flora and fauna.
User control: Compatible open-weight models can potentially run locally, giving users greater control over where their mission-generation requests are processed.
Accessibility: Local inference can avoid per-request charges from hosted AI APIs, although running a model requires suitable hardware and resources.