MappedOut AI

Community infrastructure, made visible.

Comment

MappedOut AI

A two-sided civic tech platform that routes resident complaints to the right government agencies — using plain language and AI.

💡 Inspiration

Infrastructure problems don't get fixed because they go unreported — they go unreported because the system is exhausting. Residents submit complaints and hear nothing back. They don't know if anyone read it, if it's being handled, or if it even went to the right place. That silence breeds frustration, distrust, and duplicate reports piling up in inboxes that are already overwhelmed.

On the other side, government agencies and infrastructure companies are drowning in data — sifting through emails, spreadsheets, and fragmented reports just to understand what's happening on the ground. The information exists. It's just buried.

We built MappedOut AI to fix both ends of that problem at once. We wanted to close the gap between the exhaustion of waiting and feeling ignored and the reality of swamped, disorganized data — and replace it with something transparent, efficient, and calm.

🗺️ What It Does

MappedOut AI is a two-sided civic infrastructure reporting platform connecting residents, government agencies, and infrastructure companies in real time.

Resident Side (Mobile App)

  • Residents describe their issue in plain, everyday language — no forms, no jargon
  • OpenAI categorizes the complaint and routes it to the correct agency or department
  • Before submitting, residents can see if others have already reported the same issue — including a summarized version of existing reports and a tally of how many people have flagged it. If their problem is already logged, they can simply add to the tally instead of creating a duplicate
  • All reports are stored persistently in the database — nothing gets lost, and residents are encouraged to log their cases knowing they'll be tracked
  • A greyscale Mapbox map displays complaints as colored zone overlays, color-coded by status:
    • 🔴 Red = Unresolved
    • 🔵 Blue = In Progress
    • 🟢 Green = Resolved
  • Residents can see the real status of their report at any time — genuine transparency, not empty acknowledgment

Government / Employee Side (Desktop Dashboard)

  • Officials and infrastructure teams see incoming reports organized by category and urgency
  • Some reports are automatically flagged based on their relevance to infrastructure — surfacing what matters most without manual triage
  • Each case follows a clear workflow: Problem → In Progress → Resolved, matching how teams actually operate
  • Each case includes a protocol checklist to guide resolution steps
  • Real-time sync via Supabase means both sides always reflect the latest status — no lag, no manual updates

Target users: Government agencies, infrastructure companies, and everyday residents. MappedOut AI is designed to serve all three simultaneously — as an aid to data filing, cleaning, and cutting through the noise of countless unstructured reports and emails.

🧭 How To Use It

As a Resident

  1. Open the resident app and describe your issue in plain language (e.g. "there's a huge pothole on 107 Ave that keeps getting worse")
  2. The app checks if a similar issue has already been reported in your area — if so, view the summary and add yourself to the tally instead of filing a duplicate
  3. If it's a new issue, submit your report — it will be categorized and routed automatically
  4. Track your report on the map using the color-coded zone overlays to see its real-time status

As a Government Employee / Infrastructure Team

  1. Log into the desktop dashboard to view all incoming reports
  2. Flagged infrastructure-relevant reports are surfaced automatically for priority review
  3. Open a case and move it through the workflow: Problem → In Progress → Resolved
  4. Use the protocol checklist within each case to guide your resolution steps
  5. Status updates reflect immediately on the resident-facing map — closing the transparency loop without any extra communication needed

🛠️ How We Built It

Tool Role
Base44 Full-stack app builder for both the resident and government interfaces
OpenAI API Natural language categorization, agency routing, and report summarization
Mapbox Greyscale map with dynamic zone overlay visualization
Supabase Shared real-time database connecting both sides of the platform

🧱 Challenges We Ran Into

  • Time pressure — we built and deployed the resident-facing app for the semi-finals, then built out the full government dashboard afterward — essentially shipping two complete apps in one hackathon window
  • Map design — getting the zone overlays (not pins) to render correctly by status and update in real time required significant iteration
  • AI routing accuracy — prompting OpenAI to reliably categorize vague or informal complaint language took a lot of testing and refinement
  • Duplicate report logic — designing a system that detects similar reports and surfaces a summary and tally cleanly, without losing individual context, was both a UX and data challenge

💬 If you tested either the resident app or the government dashboard during the demo and ran into any issues or unexpected behavior, please leave a comment below — your feedback is genuinely appreciated and will directly shape what we build next.

🏆 Accomplishments We're Proud Of

  • Shipped a fully deployed, two-sided platform with a live URL
  • Built duplicate report detection so residents never have to wonder if they're repeating themselves — they can just add to the count
  • Successfully implemented AI-driven complaint routing from plain-language input
  • Built a government dashboard with real case management, auto-flagging, and protocol tracking — not just a front end
  • The map visualization works: color-coded zone overlays updating by report status in real time
  • Replaced the feeling of "did anyone even read this?" with genuine, visible transparency on both sides

📚 What We Learned

  • How to scope and prioritize ruthlessly under time constraints
  • How to design AI prompts that reliably categorize informal, real-world language
  • The value of a shared database architecture when building multi-user, multi-role platforms
  • That the biggest UX problem in civic tech isn't the interface — it's the silence after submission
  • How to design for trust: residents don't need perfection, they need to feel seen and informed

🚀 What's Next for MappedOut AI

  • Add resident push notifications when their report status changes
  • Sending emails to residents whenever status on reports change
  • Expand AI routing to handle multilingual input to serve more communities
  • Build analytics for government dashboards to identify infrastructure trends and hotspots by zone over time
  • Improve report summarization so employees can scan large volumes of similar complaints instantly without reading each one
  • Pilot with real municipal governments and infrastructure companies

Built at the Black Youth AI Hackathon 2026 (SheBuilds) — Challenge Statement 3 Team: Uche, Nsikan, Taré

Built With

Share this project:

Updates