MonsoonMap
Start Walk
Hacktoberfest x DEV Week 1 • “Touch Grass” Challenge

Walk your streets. Expose the problems.
Make the evidence count.

Ordinary neighborhood walks reveal critical environmental issues—blocked storm drains, submerged curb ramps, and broken footpaths after rainfall. MonsoonMap uses open-weight vision AI to turn what you discover outdoors into structured, human-verified civic evidence.

Designed for pocket carry. Put your phone away while walking outdoors.
The Observation Workflow

From sidewalk puddle to verified evidence

No endless screen-scrolling or gamified badges. Just clear, disciplined observation that respects your time outdoors.

01

Take an Outdoor Walk

Choose a 10, 20, or 30-minute route. Focus on rainfall waterlogging, drainage grates, or footpath safety. Put your phone in your pocket.

Target: 20m rain survey
02

Capture Visible Condition

When you spot an obstacle—like standing water blocking a ramp—snap a quick photo and jot a 1-sentence note from a safe vantage point.

Photo + Field note
03

Open-Weight AI Analysis

A local vision-capable Gemma model extracts observable physical factors, isolates facts from assumptions, and suggests follow-up checks.

Gemma 3 Vision
04

Human Review & Report

You review the findings, adjust any classifications, and approve the report. Export clean PDF evidence for civic assemblies or community boards.

Verified civic record
OpenStreetMap Integration

Visualizing recurring neighborhood choke points

See where waterlogging recurs, which intersections have blocked drains, and where damaged pavers pose wet-weather slip risks. Filter by category, inspect photo evidence, and track neighborhood conditions over time.

Active Field IndicatorsLive DB Active
Waterlogging & Standing WaterHigh Priority
Blocked & Choked DrainageInlet Obstruction
Damaged Footpath PavementsTrip Hazard

Gemma 3 Open Weights

We connect directly to open-weight vision models (such as Gemma 3) running on your own machine via Ollama. No proprietary closed APIs; zero recurring vendor lock-in.

Zero Cloud Surveillance

Photos of your neighborhood and daily walking paths stay in your browser and local persistence. We never upload your images to third-party commercial clouds.

Disciplined Evidence

The AI strictly distinguishes visible physical facts from assumptions. You retain 100% human editorial authority over every final report.

Ready to step outside and observe?

Grab an umbrella, put your phone in your pocket, and walk your neighborhood block for 15 minutes. Notice what the rain reveals.