The platform, up close

Three modules. An assistant. One closed loop.

Horizon anticipates. Tower decides. Hangar learns. Each phase feeds the next prediction. One platform that closes the loop most operations leave open: between disruption, decision, and the lessons that should have made the next one easier.

Horizon
Tower
Hangar
+ Wingman the AI assistant threading through them all
Orchestration

A layer that connects what you already run.

Your operation already has the data. It lives in Movement Control, Crew Management, MRO, Flight Planning, half a dozen systems built for half a dozen jobs. X360 doesn't replace any of them. It reads across them through a single integration layer, decides with all of them in scope, and writes results back where they belong.

Layer 03 Output What X360 sends back.
X360 Writeback
Decisions and updates pushed to source systems
API Gateway
Programmatic access for downstream consumers
Alerts & Notifications
Real-time signals to humans and channels
Layer 02 stratosX Core Platform The decision engine.
Delay Prediction Models
ML ensemble · 72-hour horizon
Analytics Engine
Historical & real-time
Simulation Engine
What-if scenario modeling
Recovery Optimization
Multi-agent solver
stratosX Integration Layer
Layer 01 Your airline systems What you already run.
Movement Control
REST API
Crew Management
AIDX / XML
MRO / Maintenance
REST API
Flight Planning
ARINC 633
DCS / Check-in
HTTP Push
Weather Services
REST API
Schedule Mgmt
CSV
Airport Ops
REST API
Where the work happens

One module per phase. The assistant runs alongside.

Each phase of the loop is its own surface, purpose-built for the work operators actually do at that moment. Wingman threads through all of them.

Horizon · Pre-Ops

Horizon

See pressure building before it becomes a disruption.

Network-level delay predictions, probability scores, and proactive planning. Continuous monitoring. Assessment starts before disruption lands.

  • Delay probability tracking across the network
  • Weather and resource pressure alerts
  • Multi-Flight What-If Simulator
  • Branching scenarios at 30, 60, 90-minute horizons
AI / Analytical tools applied
Multi-Horizon Prediction Random Forest Neural Networks Propagation Modeling Probability Distribution
Tower · Day of Ops

Tower

Manage by exception — real-time alerting with integrated, cost-aware recovery.

A multi-agent system where specialized agents for aircraft, crew, and passenger domains run in parallel, exchange proposals, and converge on cost-ranked recovery options.

  • AI-driven proactive alerts before thresholds breach
  • Aircraft, crew, and passenger constraints evaluated together
  • Cost-ranked recovery options in minutes
  • End-to-end disruption workflow: detect to resolve
AI / Analytical tools applied
Approximate Search Stochasticity-Aware Multi-Agent Optimization Adaptive Learning
Hangar · Post-Ops

Hangar

Every resolved event makes the next one faster.

Resilience analytics, root cause analysis, and cost attribution. Outputs feed back into Horizon's predictions and Tower's recovery, closing the loop.

  • Per-event cost visibility: crew, pax, ops
  • Root cause patterns and KPI performance trends
  • Resilience framework: Pressure Index, Propagation, Time-to-Recover
  • Resilience Quadrant: executive-ready station classification
AI / Analytical tools applied
Descriptive Analytics Diagnostic Analytics Prescriptive Analytics Closed-Loop Learning
Wingman · AI Assistant

Wingman

One question replaces five dashboard clicks.

Embedded in your operations chat. No new tool to learn. 46 specialized tools, every answer grounded in live data. Tailored to your delay codes, your ground times, your routes.

  • Real-time flight status, rotation, and weather intelligence
  • ML delay predictions and root cause on demand
  • OTP metrics, station rankings, network health snapshots
  • Export to CSV, Excel, or PDF
AI / Analytical tools applied
Natural Language Processing Retrieval-Augmented Generation Contextual Memory Multi-User Collaboration
How it fits

Decisions flow through X360. Operators always sit in the middle.

01 · Propose
Our solvers, or yours.
  • X360 ships with its own solvers
  • Got a crew solver? Tower wraps it
  • PSS flows? Aircraft actions? Plug them in
02 · Approve
Human gate, always.
  • Nothing commits without operator approval
  • Agents propose, controller decides
  • Augmented decision-making, not automation
03 · Push back
Into your stack.
  • Writeback into Movement Control, Crew Mgmt, PSS
  • Through your sandboxes, not around them
  • The platform doesn't sit beside your stack. It acts on it
A real event, end to end

From disruption to recovery, in 5 minutes.

Flight 638, a winter charter from a regional origin to a leisure destination. A mechanical issue at the gate compounds into a crew legality problem and a passenger compensation cliff. Here is what happens with X360 in the loop, and how that compares to today.

Stage 1 Detect
Horizon

Flight 638 mechanical at origin. Downstream cascade mapped in 90 seconds.

$85K+
cascade exposure identified
Today: 15-min awareness gap, no cost estimate.
Stage 2 Assess
Horizon & Tower

Compound: crew duty exceeded, cabin crew unfit, the 3-hour compensation threshold looming. All desks assess simultaneously.

5 min
full assessment
Today: sequential desk-by-desk takes 60 to 90 min.
Stage 3 Decide
Tower

Three recovery options generated and ranked by total network cost. The cheapest direct option is not the cheapest once cascade is in.

$49K
saved vs the cheapest direct option
Today: first workable solution, with about 40% of the cost picture.
Stage 4 Learn
Hangar

Pattern captured. Recovery template stored. Levers measured. The next compound disruption at this station starts smarter.

15%
projected Y1 cost reduction
Today: data in silos, no learning loop.
The X360 difference
90 seconds
vs 15 min today
to first awareness, with the cascade already mapped
5 min
vs 60–90 min today
to ranked recovery options, with full cost in scope
Show capability detail and the three recovery options Hide detail
Stage
X360 capability
Operational impact
Est. savings
Soft benefit
Detect
Auto-detection & cascade mapping
OCC aware in 90s; downstream flights and rotation flagged instantly
~$5–10K / event
Common operating picture from minute one
Assess
Simultaneous all-desk assessment
Full picture in 5 min vs 60–90 min sequential
~$8–15K / event
Crew, pax, aircraft desks see the same data at the same time
Assess
Threshold & legality countdowns
Compensation thresholds, crew legality windows flagged in advance
~$5–12K / event
Decisions made before thresholds, not after
Decide
Option ranking by network cost
Best option identified in minutes; cascade impact in scope
$49K (option A vs C, below)
Team sees the rationale behind the recommendation
Decide
Full cost visibility
Aircraft + crew + passenger + compensation rolled into each option
Hidden cost surprises avoided
Finance and ops working from the same numbers
Learn
Template library & lever measurement
Recovery patterns reusable; the 7 levers measured per event
15% Y1 target
Knowledge retained through staff turnover
The three recovery options Tower ranked Network cost, not direct cost. Compensation under the threshold = $127K avoided.
A
Crew rest at destination
$97K
Cheapest direct cost. But 3 downstream flights affected. Longest delay. Full cascade.
B
Overnight at origin
$142K
Moderate cost. Passengers at origin hotel. Return crew swapped. One downstream flight affected.
C
Ferry from regional base
$168K
Highest direct cost, but zero cascade. 578 passengers protected. Reserve crew ferried. Recommended.

$49K difference between waiting (Option A) and the optimal swap (Option C). Without full cost visibility, the OCC defaults to the first workable solution.

What this means for you

Same platform. Different day, depending where you sit.

Everything you've just seen, viewed through the role you actually fill. The features you'll touch first. The ones that matter most. The ones that quietly do their job in the background.

Horizon

Anticipate
  • Network pressure monitor
  • Pre-shift simulation
  • Disruption forecasting (7d)
  • Prepositioning recommendations

Tower + Wingman

Decide
  • Cascade detection
  • Multi-agent recovery
  • Cost-ranked options
  • Wingman context surface

Hangar

Learn
  • Decision economics
  • Recovery retrospectives
  • Resilience benchmarking
  • Pattern library
For everyone
A complete recovery system, with the parts that matter most surfaced for the role you fill.