Fligtht Connection Risk
Connection Risk API
Predict missed connections before they happen.
Real‑time and historical connection‑risk scoring with gate‑to‑gate walking times, immigration/security process, and airline/terminal intelligence—delivered as a developer‑friendly API.
Worldwide Coverage
Reliable and accurate flight delay data from the Top 20 major airports worldwide.
Live operational signals + historical performance + typical terminal transfers.
Connection Risk
Fast & Reliable
Ultra-fast API infrastructure with the popular formats and simple access by API Key.
Reduce
Reduce rebooking cost by prioritizing at‑risk connections before they break.
Protect
Protect customer experience with proactive messages and smarter re‑shopping.
Tune
Tune MCT policy using observed transfer performance by hub/terminal.
We return a miss connection probability (0–100%) plus the top drivers behind the score — not a black box.
How it works (Steps)
Inputs — hub (IATA), arrival flight, departure flight.
Engine — combines schedule buffer, carrier/airport reliability, deplane time variance, walk/transfer times, typical immigration/security and baggage pick-up/recheck logic by connection type.
Output — miss connection risk, average need time and available time.
Features
Top 20 Hubs ( IST, AMS, CDG, LHR, FRA, ORD, ATL, DXB, JFK, DFW, PVG, ICN, KUL, MUC, MAD, YYZ, FCO, BKK, LAX, IAH ) to be expanded
Gate/Terminal-aware walking times, people movers, monorails
Schengen, Airline partners, Local citizen considerations
Historical Flight Delays: Daily Updates 👍
Live Flight Delays: 15-minute Updates 👍
| Static MCT | Connection Risk API |
|---|---|
| One size fits all | Per‑connection probability with drivers |
| No live context | Blends live delay, gates/terminals, transfer times |
| Opaque rationale | Transparent factor weights |
| Hard to tune | Versioned models + changelog |
Why this beats static MCT
Try the API
Enter 2 connecting flights at the same hub. And mark if you are a local citizen (i.e. same country) as the hub. Demo limited to 5 requests per minute per browser.
{
"demo": false,
"miss_risk_live": {"live":"46%", "live_plus5":"36%",
"live_plus10":"31%", "live_plus15":"25%",
"live_plus30":"14%"},
"totals_live":{"avg_need_live":138, "avg_avail_live":137},
"times_live":{"deplane":"6%", "walk":"27%", "immigration":"18%",
"baggage":"25%", "security":"13%", "boarding":"11%"}
}
Use cases
Airlines/alliances: IRROPs triage; protected connection logic; seat/gate prioritization.
OTAs/metasearch: Sort by reliability; nudge to safer buffers; dynamic guarantees.
Airports: Calibrate MCT; plan signage & shuttles; inform passenger comms.
Insurers/cards: Underwrite trip interruption; proactive assistance triggers.
Trust, coverage & performance
Data sources: Live operational signals + historical performance + typical gate transfers.
Latency: P50 ~200 ms (target); versioned models with changelog.
Throttling: Token-bucket RPM with 5-minute refill (see avg/burst in pricing)
Privacy: No PII required. HTTPS only. Keys scoped per environment.
Security: Key‑based auth, optional IP allowlist. Audit logs by request_id.
Reliability: 99.9% target; idempotent retries recommended for 5xx/timeout.
No hidden costs. No surprises.
Unit rules: 1 historical call = 1 unit. 1 live call = 4 units (Pro & Enterprise Only).
FAQ
-
Hub, arrival flight, departure flight, and local citizen.
-
Probability of misconnection (0-100%).
-
Live signals are polled every 15 minutes. Historical data stores 4 weeks rolling.
-
Each historical call consumes 1 unit; each live call consumes 4 units. Your monthly allowance is a single pool you can mix. (Pro & Enterprise Only)
-
We use a token-bucket over 5 minutes so you get a safe average rate and room to burst temporarily.
-
Where available, responses include fuller context and details (e.g., all components, approximate values in minutes). Starter is redacted; Core unredacts historical; Enterprise unredacts live.
-
Yes—use overage at the listed per‑1k unit rate. Enterprise can pre‑commit for deeper discounts.
-
We log minimal request metadata for debugging and rate limiting. No PII is required or stored by default
Ready to get started?
Create an account instantly and start using the API.