NorthSky AERO / Platform
Built for airports.
Powered by aviation AI.
One platform across passenger experience, terminal, apron, runway, airside, and turnaround. One tenancy. One audit trail. Deploying across 130+ AAI airports.
What makes us a platform,
not a collection of tools.
PILLAR 01
Built for airports.
Aviation-first products. Aviation-trained models. ICAO Doc 9184 · DGCA quarterly CSS · ACI ASQ by design.
Platform guarantees →
PILLAR 02
One customer record.
The passenger at the gate, the complainant in the lounge, and the apron exception are one record end-to-end. Modular licensing — start with one product, add the rest on the same tenancy.
See the products →
PILLAR 03
Responsible by design.
Human-in-the-loop on sensitive categories. Audit trail on every automated decision. In-region inference. No customer data trains shared models.
Read AI governance →
What you can rely on.
01 · TENANCY
One login.
One audit trail.
One audit trail.
Add a zone, you don't start a new implementation. The same access controls, the same admin surfaces, the same support contract — across every product you've turned on. Tail-attributed, audit-hashed, immutable.
One tenancy · seven productsLive · AAI
02 · EDGE
Inference where
the signal is.
the signal is.
Sub-second inference at the stand, the pole, or the FOD vehicle. Bandwidth-light uplink — structured events, not raw streams. The stand keeps watching if the centre drops; the centre catches up when it returns.
Edge AI · tower-outage tolerantLive
03 · RESIDENCY
Data residency
by region.
by region.
Compute, storage, and inference scoped to your region of choice. India for AAI and government customers (DPDP-aligned, DGCA data-sovereignty posture). Other regions on request. Customer data does not train shared models.
India · others on requestLive
04 · INTEGRATION
Plugs into
what's already there.
what's already there.
A-CDM, AODB, DCS, BHS, ATC log, METAR, WhatsApp, SMS, email — read in, normalised, time-aligned on one platform clock. One taxonomy across every product, ready for the ATOMS bus.
Standards-first ingestionLive
05 · GOVERNANCE
Every decision
logged.
logged.
Every classification, every override, every model version — recorded with confidence, actor, and timestamp. Human-in-the-loop on sensitive categories. The audit log writes itself, in a format the regulator already reads.
AI governance · HITL on sensitiveLive
06 · ROLLOUT
Pilot template.
Network rollout.
Network rollout.
Prove the operation end-to-end at one site, then replicate the template across the network — surveyor app, kiosks, QR, dashboards, regulator export, all configured by code. Weeks per site, not quarters.
2 live · 130+ on the rolloutJul–Aug 2026
These are platform-level commitments — they apply to every product on the suite. See the seven products and their AI families → for what each one does on top of them.
The standards
aviation runs to.
◆ ICAO
Doc 9184Airport service-excellence framework — informs reporting structures across feedback and complaints.
◆ DGCA · INDIA
Quarterly CSSCustomer-satisfaction submissions in DGCA format, on DGCA cadence — auto-generated for AAI today.
◆ DATA
DPDP Act 2023India data hosting, named data fiduciary, consent management, audit trail by design.
◆ INFOSEC
ISO 27001Information security management system. In progress
◆ AI MS
ISO/IEC 42001AI management system framework. Evaluating
Full procurement detail — AI governance, residency, security, architecture — in the Trust Center →
Outcomes from the
AAI pilot airports.
DEPLOYMENT SCALE
130+
AAI airports on the rollout. Two live today; full network Jul–Aug 2026.
SURVEYOR ADOPTION
98%
Field staff on the app within the first issue cycle.
RESPONSES AUTO-CLASSIFIED
250K+
Categorised across 100+ languages; human review on sensitive categories.
DGCA REPORT · AUTO-ASSEMBLED
~2 hr
Raw responses to regulator-ready DGCA submission.
How the platform
sees the airport.
Every product in the suite uses this same four-layer flow. Signal in at the airport, inference at the edge, operating record on the platform, structured events on the brain's bus.
INPUTSLayer 01
At the airport.
Every signal that matters
Cameras · apron, runway-end, taxiway, terminal
Sensors · radar, mm-wave, weather, friction, weight-on-wheels
IoT & edge devices · stand boxes, mobile vehicle, GSE trackers
Kiosks · terminal touchscreens at the exit
Mobile apps · surveyor tablet, passenger QR & SDK
Third-party APIs · A-CDM · AODB · DCS · BHS · ATC · METAR · WhatsApp · SMS · email
EDGELayer 02
Processed at the surface.
Inference where the signal is
Edge AI inference · sub-second at the stand, pole, or vehicle
Camera + sensor fusion · the vote that confirms the event
Tower-outage tolerant · the stand stays watched if the centre drops
Bandwidth-light uplink · events, not raw streams
Multilingual NLU · runs at the edge for the kiosk & the survey app
Time-aligned · every signal on the same platform clock
PLATFORMLayer 03
The operating record.
Stored, surfaced, decided on
Operating record · tail-attributed, audit-hashed, immutable
Real-time alerts · to the radio, the console, the screen that matters
Dashboards per persona · CEO, AOCC duty, tower, airline, handler
AI decision models · sequencing, configuration, ROT, VTT, TSAT
Analytics & reports · DGCA quarterly, ACI ASQ, CEO morning brief
Audit log · every event, every classification, every override
EVENTS · ATOMSLayer 04
On the brain's bus.
The orchestration layer
Structured events · one taxonomy across every product
Tower & AOCC · the live decision view
Airline ops · per-flight scope, scoped subscription
Ground handler dispatch · per-team performance
Regulator audit · the inspection log writes itself
CEO dashboard · delivered vs declared, throughput, sustainability
See the AI at work
on your airport.
A 30-minute walkthrough on your real data — terminal layouts, current reporting cadence, the ask from your regulator. We bring the AI; you bring the operation.