TraceBack

TraceBack is a secure lost-and-found web app that helps users report, track, and recover lost items using structured workflows, privacy-first verification, and AI-assisted matching with Gemini

Comment

Inspiration

Losing personal belongings is a common real-world problem, but recovering them is often unstructured, insecure, and stressful. Most existing solutions rely on informal communication, lack privacy, and provide no clear verification during handover. TraceBack was built to explore how a structured, secure workflow could improve the lost-and-found process.

What it does

TraceBack is a full-stack lost-and-found web application that allows users to report lost or found items, track claims, and complete recoveries through a controlled and privacy-first workflow. Users can submit item reports, review potential matches, and manage claims from a centralized dashboard.

To ensure safe recovery, ownership verification is handled through an approval flow and a one-time PIN system that is used during physical handover. Sensitive contact information is only shared after a claim is approved.

How we built it

The application is built using Next.js with a modern React and TypeScript stack. Firebase is used for authentication, database storage, file uploads, and hosting. Firestore security rules enforce strict claim state transitions and access control.

Gemini is used in an experimental, AI-assisted role to explore how large language models can support matching workflows and ownership verification by analyzing item descriptions and generating verification questions.

Challenges we faced

Designing secure claim workflows and enforcing them at the database level was a key challenge. Another challenge was integrating AI in a responsible way without over-automating critical decisions such as ownership verification. The project focuses on clarity, security, and real-world usability rather than replacing human judgment.

What we learned

This project helped us understand how to design real-world systems with authentication, privacy, and security constraints, and how AI tools like Gemini can be used as supportive components within structured workflows rather than as standalone decision-makers.

Built With

  • firebase-authentication
  • firebase-hosting
  • firebase-storage
  • firestore
  • google-gemini
  • google-genkit
  • hook
  • next.js
  • react
  • tailwind-css
  • typescript
  • zod
Share this project:

Updates