The problem

Montgomery, Alabama has no modern civic reporting platform. When a resident hits a pothole or finds a broken streetlight, the only channel is a multi-step form with no community visibility, no confirmation, and no feedback loop. City teams receive unstructured duplicates with no photos, no GPS, no priority signal. Fixable problems persist for months.

The barrier is friction, not indifference. Infrastructure that could be repaired cheaply becomes an emergency because the signal never reaches the people who can act on it.

· · · · · · · · · · · · ·

NEIGHBORHOOD ISSUE RADAR

One photo. AI detects the issue. Ten seconds to submit.

A real-time civic map for Montgomery. Upload a photo — AI pre-selects the severity level and offers a one-click professional description. You select the issue type and add a title. The report appears on the shared map within one second via WebSocket, visible to every resident. Status updates in real time. The feedback loop closes.

Address search with map highlight — type any street or address worldwide and the matching street glows indigo on the map. Duplicate detection within 150 m concentrates community signal. A predictive risk engine scores infrastructure corridors 0–100. An AI insight engine surfaces the most urgent pattern in one sentence. Live NOAA weather with NWS emergency alerts. 3D city view, heatmap, voice dictation, colorblind themes, World Context via AI. All running in production, at zero cost, right now.

· · · · · · · · · · · · ·

~40 hours of active development. Live in production. Ready to scale.

GenAI Works Hackathon · World Wide Vibes · March 5–8, 2026

· · · · · · · · · · · · ·

A note before you begin

I am currently recovering from Epstein-Barr virus. Extended or sustained speech is medically contraindicated at this stage — there is no voiceover for this demo. One exception: the voice dictation feature (Web Speech API) will be demonstrated briefly, as it is a core accessibility function of the platform. That single use is the full extent of vocal input in this presentation.

The demo runs live in production, signed in with a Google account pre-configured with admin privileges.

Admin access allows deleting and moderating any report directly from the map — city staff use the same interface as residents. A real street photo will be uploaded live: AI fills every form field in under two seconds. The ✨ Suggest button will also be tested manually: without a photo, the user selects category and severity, types a title, and AI drafts a professional description on demand.

After this crawl, the interactive demo launches automatically — eleven steps across every feature, on real live data, in real time.

The 18 reports on the map are not test data. Each one is based on a real, documented civic incident from Montgomery — chronic flooding on Hayneville Road, a 100-unit abandoned apartment complex, a water main break that triggered a boil advisory for 10,000 residents, five road corridors approved for emergency resurfacing after years of visible deterioration. Sources: WSFA News, City Council records, Montgomery Open Data Portal.

Report dates are set within a five-month window (Oct 2025 – Mar 2026) so the map reflects a coherent current snapshot: open issues are genuinely unresolved, in-progress ones have city action underway, resolved ones were closed. The OVERDUE badges you will see are not decoration — they mark real problems that sat ignored for weeks. You can also explore freely at neighborhood-issue-radar.vercel.app.