Active programs
9
AI for airports, airlines, and air-traffic operations aligned to GACA's mandate and the Vision 2030 aviation strategy. We build alongside Riyadh Air, the new mega-terminals, and the legacy carriers — engineered for the operational rhythm of a kingdom becoming a global aviation hub.
Six pressures we hear from COOs, airport directors, and airline operations control on every first call.
Vision 2030 calls for 330 million passengers per year. The capacity, the IT, and the operational data fabric inherited from the previous decade do not get there without an AI-first re-platforming.
One thunderstorm over Riyadh cascades through 200 flights. Crew, slot, gate, and bag recovery is still a war-room exercise — it should be a model that proposes the recovery plan before the controller asks.
Two predictable, multi-million-passenger surges per year. Capacity, immigration, ground handling, and connecting domestic flights have to absorb the spike — generic airport planning tools miss the cultural and seasonal pattern.
Heavy maintenance is the largest controllable cost line for any carrier. Sensor data, fault histories, and maintenance manuals exist — connecting them into a model that predicts the next removal is the unsolved part.
Every passenger touchpoint has to work in Arabic and English at parity. Most airline and airport software was built English-first and bolted Arabic on later — it shows in the call-centre and the kiosk.
Airports are critical national infrastructure under the NCA's CCC. The OT and IT estates were converged before they were secured — and the threat surface keeps growing as more systems become networked.
Six disciplines, sector-tuned for hub operations, MRO, and the passenger journey.
A unified operational data layer across AODB, FIDS, BHS, RMS, and the airline ops control feeds — the substrate every other AI workload depends on.
Agents that draft the disruption recovery plan — crew, slot, gate, bag — and let the controller approve, edit, or reject. Audit-grade, with a clear rollback path.
Turnaround timing, gate-event detection, queue analytics, and ramp-safety monitoring from existing CCTV — no new sensors required.
In-Kingdom landing zones for airport and airline workloads, with the integration patterns the legacy AODB / BHS estate actually needs.
An OT-aware SOC posture for airports and carriers, aligned to NCA CCC and the IATA cyber framework. Built around the IT/OT convergence reality.
Removal forecasting and maintenance optimization across the heavy and line MRO estate. Sensor + fault + manual fusion, modelled per fleet.
Two engagements that anchor our aviation practice. Names redacted under MNDA — the operations directors know the work.
An ops-control copilot that watches the day's schedule, detects emerging disruption, and proposes a recovery plan — slots, gates, crew, bags — before the controller asks. The controller approves, edits, or rejects; the model learns from every decision. No surprise actions, no autonomous commits.
Predictive removal forecasting on a wide-body fleet, fusing sensor histories, fault codes, and maintenance manuals into a single model. The output is a 30-day removal queue the maintenance planner trusts — not a heat-map dashboard the engineer ignores.
Authorities, operators, and the integration partners we work with at the airport and carrier level.
No. We sit alongside them. The AODB stays the system of record; we add the data fabric, the AI workloads, and the operator-side surfaces that the legacy estate cannot deliver natively.
Surge-aware capacity and operations planning is built into the platform — not bolted on. Models are trained on the historical surge pattern (and the cultural calendar, not the Gregorian one).
No, by design. It proposes; the controller commits. Every recovery plan is logged, with the model's reasoning, the controller's edits, and the post-recovery outcome — for audit and continuous improvement.
Our aviation SOC posture is OT-aware from the ground up. We segment IT and OT traffic, monitor the BHS / FIDS / AODB control planes as first-class assets, and align controls to NCA CCC and IATA's cyber framework.
Yes. We model per fleet (and per engine variant where it matters) and unify the outputs into a single planner-facing queue. Wide-body and narrow-body have different signal quality — the architecture accounts for that.
One quarter to a narrow, in-production use case — most often disruption recovery or apron CV. The data fabric is the multi-quarter investment; the workloads on top ship in weeks once it is in place.
Sixty-minute working session with our Aviation lead and an ops-control specialist. Bring the operational pressure — disruption, surge, MRO, passenger experience — and we'll come back with a one-page roadmap.