Most booking channels show the itineraries that airlines have agreed to sell together. That leaves out a large space of journeys that combine low-cost and full-service carriers with no agreement between them. For travellers who put price first, that space is where the value is.
The network as a graph
The approach in our delivered work treats every airport as a node and every flight as an edge carrying fare, times, carrier type, baggage rules and connection constraints. Content from GDS, NDC and direct LCC APIs plugs into the same graph, so adding a carrier means adding edges rather than redesigning the system1.
Feasibility before optimisation
A cheap route that misses its connection is not cheap. The engine checks minimum connection times per airport and carrier, terminal changes, immigration time for international transfers, self-connect baggage requirements and transit-visa eligibility by passport, and drops infeasible edges before any ranking happens1. Only then does a multi-criteria optimisation rank the fastest, cheapest and best-value journeys.
New Delhi to San Francisco: leading OTA vs our engine
Example search from our delivered OTA-marketplace work
| Source | Option | Travel time | Cost (USD) |
|---|---|---|---|
| Leading OTA | Cheapest | 35h 30m | 856.70 |
| Our engine | Cheapest | 24h 20m | 702.44 |
| Leading OTA | Fastest | 15h 30m | 1,405.72 |
| Our engine | Fastest | 15h 30m | 1,044.54 |
Note: 22 airlines analysed, 392 virtual combinations.
For OTAs, the prize is a traveller segment that traditional channels underserve. For airlines, the same graph is a planning tool: it shows which partners and connections would open demand the network cannot capture today.