Got2Get2Work

AI-coordinated, employer-sponsored commute resilience for shift workers.

Comment

The problem

A reliable worker can still lose income when transportation fails. Shift-based employers absorb absence, overtime, and staffing disruption, while conventional ride-hail treats every commute as an isolated transaction. Coworkers may already travel toward the same worksite at compatible times, but they lack a trusted, privacy-conscious way to coordinate recurring carpools and recover from cancellations.

What Got2Get2Work does

Got2Get2Work is an Android-first carpool coordination marketplace designed to operate as an employer-sponsored workforce benefit. Workers verify accounts and worksite membership, confirm recurring schedules, choose Driver, Passenger, or Either independently for each direction, and review explainable coworker options.

Deterministic policy code enforces worksite membership, compatible time windows, active consent, blocks, seat availability, and detour limits. Both coworkers must agree before pickup and vehicle details are revealed. Cancellation recovery may surface a standing backup, but the platform never represents transportation, attendance, reimbursement, or ride completion as guaranteed.

The employer console provides worksite enrollment controls and privacy-safe aggregate outcomes. It excludes home addresses, precise location trails, private messages, plates, individual schedules, safety narratives, ratings, and attendance records. Small cohorts are suppressed.

How the business operates with AI

Gemini operates as a privacy-limited site coordinator. It receives only allowlisted aggregate facts such as enrollment, confirmed rides, cancellation activity, recovery activity, and self-reported protected shifts. Gemini returns a structured recommendation using server-owned action codes and cites the aggregate fact identifiers supporting it. The server validates the response before display.

Gemini cannot approve a worker, override a safety restriction, reveal an address, contact employees, authorize spending, or promise a ride. Deterministic matching and authorization remain authoritative. A visibly labeled deterministic fallback preserves availability when the model is unavailable.

Workers retain control of participation, schedules, requests, acceptance, and proximity sharing. Employer administrators verify roster membership and manage pilot scope. Humans remain responsible for safety, privacy, spending, financial reporting, and consequential employment decisions.

Architecture

The production architecture is API-first and modular: a typed Expo/React Native client, typed Fastify services and repositories, Firebase Authentication, Firestore, Secret Manager, Gemini, and containerized deployment on Google Cloud Run. The system uses structured errors, logging, strict CORS, dedicated runtime identities, revocation-aware session checks, tenant-scoped authorization, and transactional safety controls.

Business model and impact

Got2Get2Work is a B2B2C employer-sponsored benefit. Employers, workforce programs, or worksites would pay a platform or pilot fee; workers would not pay a monthly subscription. The first target market is South Florida shift-based organizations in healthcare, hospitality, logistics, senior care, education, and municipal workforces.

Impact is measured through verified accounts, approved members, match discovery, acceptance, confirmed carpools, cancellations, successful backup matches, participant completion responses, and self-reported protected shifts. Fictional demonstration metrics are never represented as customer results. Revenue, expenses, users, and testimonials are disclosed exactly as evidenced, including zero values and related-party revenue.

What humans do versus AI

Codex supported architecture, implementation, debugging, testing, security review, deployment, and submission preparation. Gemini's production role is narrower: transform privacy-safe aggregate facts into validated operational recommendations. Humans make customer, safety, access, spending, legal, and employment decisions.

At scale, Got2Get2Work could protect scheduled income, reduce avoidable staffing disruption, and create implementation, customer-success, trust-and-safety, and local mobility-partnership work. Those are potential outcomes, not current claims. The immediate objective is a controlled South Florida pilot that tests whether recurring coworker coordination can protect real shifts safely and measurably.

Testing

Judges can use the isolated fictional demonstration to configure a schedule, inspect an explainable match, complete mutual request and acceptance, simulate cancellation recovery, and view the employer aggregate dashboard and site-coordinator flow without entering personal data.

Built With

Share this project:

Updates