SkyPort AI: Real-time Travel Assistant

Inspiration

Modern air travel is information-heavy but insight-poor.
Passengers can see flights, airports, and delays but rarely get contextual help when they actually need it. Finding nearby hotels during a layover, understanding flight status, or deciding where to eat near an unfamiliar airport still requires switching between multiple apps.

SkyPort AI was inspired by a simple question:
What if flight tracking, local discovery, and AI assistance lived in one real-time interface?

Instead of static dashboards, SkyPort AI turns live airspace into an interactive travel companion.

What it does

SkyPort AI is a real-time, map-based travel assistant that combines live global flight tracking, AI-powered travel guidance, and nearby recommendations around active airports.

It allows users to:

Track live flights anywhere in the world
Ask AI questions about flights, airports, and travel logistics
Instantly discover nearby hotels, lounges, and dining options
Explore airspace visually through an interactive map

SkyPort AI bridges the gap between where flights are and what travelers should do next.

How we built it

SkyPort AI is designed as a lightweight, real-time web application:

Frontend: React 18, TypeScript, Tailwind CSS, Vite
Mapping: Leaflet.js for fast, interactive geospatial visualization
Flight Data: OpenSky Network API for live aircraft telemetry
AI Assistant: Google Gemini Flash for contextual reasoning
Local Discovery: Yelp Fusion API for nearby amenities

Core Systems

Live Airspace Visualization:
Real-time flight positions rendered as vector overlays on an interactive world map.
Context-Aware AI Chat:
The assistant understands user questions in relation to active flights, airports, and map position.
Proximity-Based Recommendations:
Nearby hotels, lounges, and restaurants dynamically suggested around airports using live location data.

Challenges we ran into

Real-time performance: Rendering many moving aircraft efficiently without overwhelming the browser
Context management: Ensuring AI responses stay grounded in live flight and map data
API coordination: Synchronizing OpenSky, Yelp, and Gemini responses without blocking the UI
UX clarity: Making dense aviation data understandable for non-expert users
Rate limits: Designing around public API usage constraints

Each challenge shaped SkyPort AI into a responsive, user-friendly system rather than a raw data viewer.

Accomplishments that we're proud of

A functional real-time global flight tracking map
An AI assistant that provides contextual travel guidance, not generic answers
Seamless integration of aviation data and local discovery
A clean, responsive UI optimized for exploration and clarity
A foundation that can scale to richer travel intelligence features

What we learned

Real-time data is only valuable when paired with context and guidance
Visual interfaces dramatically improve understanding of complex systems
AI feels most useful when it responds to what the user is seeing, not just what they type
Simplicity in UX is critical when working with dense geospatial information
Travel tools should reduce cognitive load, not add to it

What's next for SkyPort AI

Flight delay and disruption intelligence
Layover-aware recommendations and timelines
Multi-airport route planning
User personalization and travel preferences
Mobile-first and PWA support
Airline, weather, and NOTAM data integration

Our goal is to evolve SkyPort AI into a real-time decision layer for global travel.

One-line pitch

SkyPort AI turns live global airspace into an intelligent travel assistant combining flight tracking, AI guidance, and nearby discovery in one real-time map.