iOS
Tarmac
iOS app that turns flight delays into spontaneous city itineraries, with an AI safety layer that never lets a bad plan reach the user.
100%
violations caught
01 — Problem
A long layover is wasted time, but an AI travel planner that doesn't understand real airport logistics — security wait times, boarding buffers — could suggest a plan that makes someone miss their flight.
02 — Approach
- 01
Two-stage AI planning pipeline: an LLM generates a raw itinerary, then a deterministic safety layer checks it against real constraints before it ever reaches the user.
- 02
Risk Level scoring built on TSA wait-time models and boarding buffers ensures timing feasibility.
- 03
MapKit handles routing data so every suggested itinerary has accurate travel-time estimates.
03 — Architecture
04 — Results
100%
boarding-constraint violations caught (0% false positives)
66%
of raw model plans violated timing constraints before the safety layer
100
simulated scenarios tested
Stack
Next
Onyx→
Voice AI that autonomously manages pharmacy calls for patients with visual impairments or language barriers.