Sentry.ai reads rainfall, soil moisture, satellite imagery and terrain data across the North Eastern Region in real time, predicts where landslides are likely, and alerts district administrations, field officials and villages while there's still time to act.
Heavy rainfall, fragile terrain and unplanned hill cutting make the North Eastern Region especially prone to landslides, flash floods and slope failures — and today, most vulnerable zones are only checked after something has already gone wrong.
Reactive monitoring. Vulnerable slopes are inspected after reports come in, not watched continuously.
Manual reporting. Field information travels by phone call and word of mouth, with no shared record.
Isolated villages. Road blockages can cut off remote communities for days before help arrives.
No predictive layer. Authorities rarely know a zone is turning dangerous until it already has.
Sentry.ai pulls together the data sources that already describe a slope's risk, scores that risk continuously, and gets the result to the people who need to act on it — before the next heavy spell of rain.
Rainfall patterns, soil moisture sensors, satellite imagery, terrain and slope data, and historical landslide records feed the same model.
Models trained on terrain and weather history flag high-risk zones and estimate the likelihood of a landslide event.
District administrations, disaster management authorities and local communities are notified as risk changes, not after the fact.
Vulnerable roads, villages and infrastructure are plotted on a live map, so risk is something you can see, not just read.
Anyone on the ground can upload geo-tagged photos or videos of cracks, slope movement or blocked roads in seconds.
Notifications go out in local languages, and the field app keeps working — and syncs later — where connectivity drops.
The same pipeline runs continuously, sector by sector, across the monitored region.
Live and historical data is pulled in from every available source for a given slope.
ML models trained on regional terrain and weather patterns rank each zone by risk severity.
Warnings reach the right people through the right channel, prioritised by severity.
One screen for risk severity, road connectivity, weather-linked forecasts and where the response effort should go first.
Sentry.ai is under active development for SIH. At a high level, it's structured in three layers, plus the connections that keep it fed with data.
Ingests IMD weather APIs, satellite feeds, ground sensor readings and citizen-submitted geo-tagged media into a single stream.
Scores each monitored zone continuously and turns that score into the heatmaps and forecasts shown on the dashboard.
Pushes multilingual SMS and app alerts, and runs the mobile field-reporting app with offline capture and background sync.
A cloud-based core with local caching on the field app, so reporting and viewing risk keeps working through connectivity drops.
We're building this for the Smart India Hackathon. Leave your details and we'll reach out when the field app and dashboard are ready to try.