{"id":6178,"date":"2026-10-20T04:12:10","date_gmt":"2026-10-20T07:12:10","guid":{"rendered":"https:\/\/tucumandevelopers.com\/index.php\/2026\/10\/20\/slow-and-steady\/"},"modified":"2026-10-20T04:12:10","modified_gmt":"2026-10-20T07:12:10","slug":"slow-and-steady","status":"publish","type":"post","link":"https:\/\/tucumandevelopers.com\/index.php\/2026\/10\/20\/slow-and-steady\/","title":{"rendered":"Slow and Steady"},"content":{"rendered":"
\n
<\/header>\n
\n

Hacktoberfest Open-Source AI Challenge Week 3: Slow and Steady Submission \ud83e\udda5<\/p>\n

\n

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 3: Slow and Steady<\/a><\/em><\/p>\n

<\/a> What I Built <\/h2>\n

TouchGrass AI is a purpose-built micro-expedition application designed around one strict rule:
The screen should be the shortest part of the experience.
Developers, students, and desk workers can spend most of their day in front of screens, and even short breaks can turn into more screen time. TouchGrass AI breaks that loop by turning a short break into a simple outdoor mission.
The experience is intentionally:
Choose \u2192 Get a quest \u2192 Pocket the phone \u2192 Go outside \u2192 Observe \u2192 Return \u2192 Verify
The user selects an available time budget \u2014 5, 15, 30, or 60 minutes \u2014 and an environment such as a park, backyard, urban area, trail, or waterfront.
TouchGrass AI then generates a sensory outdoor micro-quest with physical activities such as observing nature, listening to birds, feeling tree bark, or exploring the surroundings.
Pocket Mode then dims the screen, plays a departure chime, and tells the user to put the phone away and step outside.
After the activity, the user captures exactly one field photo. A local MobileNetV4 ONNX model runs directly on the CPU to classify the observation and distinguish nature\/outdoor findings from obvious indoor technology such as monitors and keyboards.
The activity and outdoor minutes are recorded in a private local SQLite journal.
The goal is not to keep users inside the application.
The goal is to make them leave it.<\/p>\n

<\/a> Demo <\/h2>\n

<\/a> Code <\/h2>\n
\n
\n

gokulkakde <\/a> \/ touchgrass <\/a> <\/h2>\n

TouchGrass AI is a local-first AI app that turns screen breaks into short outdoor micro-adventures. It uses open-weight AI locally to create outdoor quests and verify field observations\u2014helping users spend less time on screens and more time outside. <\/h3>\n<\/p><\/div>\n<\/p><\/div>\n

<\/a> How I Built It <\/h2>\n

Product Architecture<\/strong>
User at Desk
\u2193
Short Screen Interaction
\u2193
Select Time + Outdoor Environment
\u2193
Quest Engine
\u2193
Sensory Outdoor Micro-Quest
\u2193
Pocket Mode
\u2193
Phone Goes Away
\u2193
REAL-WORLD OUTDOOR ACTIVITY
\u2193
One Field Photo \/ Observation
\u2193
Local ONNX AI Inference
\u2193
Nature \/ Outdoor \/ Indoor Classification
\u2193
Private Local SQLite Journal<\/p>\n

Technology Stack<\/strong>
AI<\/p>\n