HomePlatformSolutionsResourcesCustomers
Case Study · Airlines · Mexico City

AeroMexico improves digital
ticketing performance with Applicare.

AeroMexico operates 340 microservices supporting flight operations, reservations, digital ticketing, and loyalty programmes across AWS and on-premises infrastructure. Applicare reduced their MTTR from 4.5 hours to 11 minutes and transformed their digital customer experience.

AirlinesDigital TicketingAWS + On-PremisesArcIn AI340 Microservices
4.5h → 11min
MTTR reduction
80%
On-call page reduction
340
Services in entity graph
$180k/mo
Estimated savings from faster MTTR

Running an airline on microservices

Modern airline operations are extraordinarily complex from a software perspective. AeroMexico's digital platform handles flight reservations, real-time seat inventory, baggage tracking, loyalty point calculations, payment processing, and digital ticketing — all simultaneously, all with zero tolerance for downtime during peak travel periods.

Their 340 microservices span AWS infrastructure and on-premises data centres, with hundreds of inter-service dependencies. When something went wrong — and in a system this complex, something always eventually went wrong — the diagnosis process was a war room exercise involving 8+ engineers, 5 separate monitoring dashboards, and an average of 4.5 hours before root cause was identified.

During peak travel periods like Christmas and Semana Santa, a 4.5-hour outage on the digital ticketing platform could affect 50,000+ passengers and generate significant revenue impact. The pressure to resolve incidents faster was existential.

Building the causal entity graph for 340 services

Applicare's entity graph auto-discovered all 340 microservices and their dependencies within the first 12 hours of deployment. No manual CMDB configuration. No infrastructure-as-code parsing. The graph identified relationships between services that the team hadn't formally documented — including several circular dependency risks that were immediately addressed.

With the complete causal graph in place, ArcIn could traverse the full dependency chain when diagnosing incidents. A slowdown in digital ticketing could be immediately correlated with the inventory availability service, which depended on the reservation database, which was experiencing connection pool pressure from a concurrent loyalty points calculation batch job.

4.5h → 11min
MTTR — across all 340 services
Live — AeroMexico transaction monitoring dashboard

The digital ticketing performance breakthrough

Within the first month, ArcIn identified a performance pattern in AeroMexico's digital ticketing flow that had been causing intermittent slowdowns for 18 months. The ticketing service was performing N+1 queries against the seat inventory database — executing one query per available seat class rather than batching — a pattern that only manifested under load during busy booking periods.

The fix was straightforward once identified: a single query change reduced ticketing API response times by 340ms at peak load. But identifying it had taken Applicare 47 seconds; the team had been chasing the symptom for 18 months.

  • Digital ticketing API p99: 2.1s → 380ms at peak load
  • Seat inventory queries per booking: 47 → 1
  • Booking abandonment rate during peak: reduced by 23%
  • Payment gateway timeout rate: reduced by 91%

The war room that never happened

Three months after deployment, AeroMexico experienced a major incident during a peak booking period. A database configuration change had inadvertently reduced the connection pool size for the reservation service. Within 60 seconds, checkout was degrading for all users.

ArcIn identified the root cause in 47 seconds and surfaced it directly to the on-call engineer's Slack. IntelliTune automatically restored the connection pool configuration in 380ms. Total user-facing impact: 4 minutes. Total engineer involvement: one person, one Slack message. No war room. No bridge call.

Before Applicare, diagnosing a p99 regression meant a war room with 8 engineers, 5 dashboards, and 4.5 hours of manual correlation. Now ArcIn tells us the root cause and the fix in under 60 seconds. Every time.

4.5h → 11min
MTTR reduction across 340 services
At a glance
CompanyAeroMexico
IndustryAirlines / Aviation
InfrastructureAWS + On-premises hybrid
Services340 microservices
Challenge4.5h MTTR, ticketing performance
Outcome11min MTTR, 340ms ticketing improvement
Get similar results → ← All case studies
Want results like AeroMexico?
See Applicare on your environment in 30 minutes.
Book a live demo →