CrewConnect

problem statement Today, hiring a licensed gig worker is messy and slow. • Agencies act as middlemen, passing calls from one person to another. By the time the job reaches the worker, hours are lost and their pay is cut down. • Gig platforms aren’t better: they force workers to bid against each other, which rewards the cheapest, not the most qualified.

The result? Delays, wasted travel, lower wages, and weak compliance.

CrewConnect replaces this broken system with instant, fair, explainable offers. • A client types their job in plain English. • Our AI scoping tool turns that into structured tasks. • A matching algorithm checks skills, licences, distance, pay, reliability, and fairness. • Nearby licensed workers get instant expiring offers, with backups ready if needed. • The first to accept wins — booked atomically with no double-assignments.

The result is minutes to staff a job, 25–40% less travel per worker, and compliance enforced by design.

⸻

💡 What Inspired Us

We looked at the cities of the future: • Growing populations demand faster, fairer infrastructure. • Climate change makes efficiency and sustainability urgent. • Workers need reliable income without wasting hours bidding or traveling across town.

We asked one question: How might we connect licensed local workers to jobs in minutes — cutting out middlemen, reducing travel, and keeping things fair?

⸻

📚 What We Learned • Explainability builds trust — showing “Why this worker was chosen” reduces disputes. • Sustainability and speed align — prioritising nearby workers makes jobs faster and greener. • Fairness needs balance — gentle nudges prevent overloading a few workers while keeping strong performers rewarded. • Atomic booking matters — first-accept-wins and backups auto-promoting are more practical than overcomplicated solvers.

⸻

🛠 How We Built It • Frontend: AI chat intake (Groq) turns free text into structured tasks. • Backend: Supabase + a custom matching algorithm. • Algorithm: Scores workers on: • Skills/licences ✅ • Distance & ETA 🌍 • Pay fit 💵 • Reliability & reputation ⭐ • Fairness (hours balanced) ⚖️ • Cancellation risk ⏱️ • Assignment: • Urgent → Instant expiring offers • Planned/Flexible → RSVP + confirm-by windows • Atomic booking ensures no double assignments.

\text{Cost}{onsite} = \text{rate} \times h + \lambda \cdot \text{travel} + \rho \cdot p{cancel} \cdot h - \nu \cdot \text{rep}

⸻

🚧 Challenges We Faced • Turning messy free text into safe, licensed task graphs. • Building realistic demo data for workers, licences, and schedules. • Preventing race conditions when multiple workers accept at once. • Balancing fairness rules without discouraging reliable workers.

⸻

✅ Outcome

CrewConnect shows that cities can move faster, fairer, and greener: • Jobs staffed in seconds (demo ~14s). • 25–40% fewer travel km per job. • Compliance gates built-in: licences, wages, and age always checked. • Fairer distribution of hours across workers.

👉 A practical, explainable staffing system — no middlemen, no bidding, no wasted time.