Community reports arrive
Structured prompts capture location area, category, urgency, and short evidence notes.
Citizen safety intelligence for the Sahel
DANAIA
A low-connectivity field intelligence layer that turns community reports into moderated public alerts and NGO-ready weekly briefs.
WhatsApp intakeTelegram intakeHuman-reviewedLow-resolution map
Live pilot loopIntake
127signals this week
Community reports arrive
Structured prompts capture location area, category, urgency, and short evidence notes.
AnimatingIntake
Start narrow, earn trust
The first version should prove signal quality before scale.
DANAIA begins where adoption already exists: WhatsApp and Telegram groups. Every report passes through moderation before anything becomes public.
Claims are clustered and reviewed
Duplicates, sensitive claims, and low-confidence reports stay in a protected review queue.
Only safe area-level alerts publish
The public layer shows zone status without exposing exact source location or movement.
NGOs receive a weekly brief
Partners get confirmed patterns, unresolved signals, and caveats for field coordination.
Product loop
From noisy community messages to safer operating visibility.
The moderation console is the product core. It protects sources, filters rumors, and creates a measured public view that NGOs can act on.
WhatsApp report82% match
Road access reported blocked near a market corridor. Two related reports in the same zone.
Approve zone alertAsk follow-up
Telegram reportNeeds review
Water point status changed. One validator confirms, second confirmation pending.
HoldRequest validator
Sensitive reportPrivate
High-risk security claim is retained for private review and excluded from the public map.
EscalateRedact
Low-resolution public map
Public alerts are generalized by zone, category, confidence, and freshness.
Route, market, health point, and water availability changes that survived review.
Confirmed, unresolved, duplicated, and withheld signals separated clearly.
Plain-language summary for field teams and partner coordination calls.
Trust and safety
The safety model is the moat.
In this category, trust is the system architecture: fewer claims, stronger review, safer publication.
Publish less than you know.
The public layer should never expose exact source identity, precise locations, or raw sensitive claims.
Make uncertainty visible.
Every alert carries a state: confirmed, developing, unresolved, or withheld for safety reasons.
Human review before automation.
Automation clusters and prioritizes reports, while moderators decide what becomes public.
90-day pilot
A revolutionary product starts with a disciplined pilot.
The first milestone is a working trust loop that proves people report, moderators can verify, and partners find the digest useful.
WhatsApp and Telegram intake, structured prompts, language variants, and simple rate controls.
Human review for duplicate reports, safety risk, confidence level, and publication status.
Area-level public alerts that are useful without revealing reporter identity or precise trails.
A field-ready brief that separates confirmed changes from unresolved or withheld signals.
Why this can spread
Distribution follows existing trust channels.
DANAIA should enter through the channels that already carry local information and become more valuable as trusted reviewers participate.
Community groups
Start where reporting behavior already exists instead of asking people to install a new app first.
Local validators
Trained reviewers turn noisy local signals into safer, more credible operational visibility.
NGO and business partners
Partners fund the system when the digest and map help teams plan around changing field conditions.
Build now
Make the first safety signal trustworthy before making it loud.
DANAIA can become the Sahel field intelligence layer by starting with one clear promise: collect carefully, verify locally, publish safely.