Flutter MVP: AI Navigation Alerts

I’m ready to move LiveRoutesSA from concept to a usable MVP and need the full Flutter build for Phase 1. The core of this release is a cross-platform (Android and iOS) navigation experience that overlays live road-safety intelligence on normal turn-by-turn guidance.

Data & integrations
– Bring in three live feeds: government traffic data, community reports and weather services. I’m open to whichever APIs, aggregators or scraping techniques you believe are most reliable for South Africa, as long as usage terms allow commercial deployment.
– Combine those feeds through lightweight AI logic (simple weighting, confidence scoring or similar) so the app surfaces only the most relevant events along the active route.

User experience
– Map screen with current position, route polyline and colour-coded markers for high-accident zones, hijack hotspots, potholes, construction and severe-weather hazards such as dense fog, mist, snow or flooding.
– Instant on-screen banners and Firebase (or equivalent) push notifications when a new hazard appears within a configurable radius or the next 5 km of travel.
– Tapping an alert should reveal a short detail card showing source, time stamped confidence score and suggested action (slow down, reroute, etc.).

Scope of work
• Set up the Flutter project, state management and responsive UI covering the above screens.
• Integrate selected data sources, schedule background refreshes and handle rate limits gracefully.
• Implement the AI merging layer in Dart (no heavy model training yet—rules or heuristics are fine).
• Wire push notifications and local alert logic.
• Package Android & iOS builds plus concise README and code comments for hand-off.

Acceptance criteria
1. Launching the app anywhere in South Africa shows the map, current GPS position and at least one live hazard marker pulled from the data feeds.
2. Starting a sample route triggers correct on-screen and push alerts whenever a simulated hazard lies ahead.
3. App runs without crashes, passes Flutter analyze, and builds cleanly for both stores.

If you can also suggest scalable back-end or analytics tooling, note it in your proposal; otherwise the focus remains the Flutter client. Show More