United Airlines strengthens operational resilience with AWS Unified Operations

Every minute of system downtime can disrupt passenger journeys and on-time performance at United Airlines. The airline operates a global technology environment spanning 2,100 applications and more than 4,300 microservices, supporting billions of digital channel interactions each day. In this environment, a 15-minute outage in the reservation system or airline operations system can affect hundreds of travelers, disrupt airport operations, and quickly cascade into flight delays across the network.

To safeguard on-time performance, United Airlines sought to strengthen resilience across its most business-critical AWS workloads, including reservation management, airport operations, fleet maintenance, and weight and balance systems. The airline required a proactive operating model that automates incident detection and escalation, accelerates expert engagement, and drives continuous improvement. “We needed a team that understood the business context of our applications, had deep knowledge of our architecture, and could help us reduce downtime across critical systems,” says Ninos Gabriel, managing director, Enterprise Architecture, Strategic Platform Engineering at United Airlines.

United Airlines engaged AWS Unified Operations to establish a proactive, integrated operating model. The service brings together designated AWS specialists, AI-driven insights, and continuous monitoring to reduce operational risk across critical airline systems and accelerate incident response. A dedicated team—including an AWS Technical Account Manager, Domain Specialist Engineers, and a Senior Billing and Account Specialist—provides guidance across resilience, observability, operations, and cost optimization. Gabriel says, “AWS Unified Operations gives us rapid access to experts across our AWS technology stack.”

United Airlines completed a 12-week onboarding program with AWS Unified Operations to deploy the new model. In the first two weeks, AWS integrated its support case system with the airline’s IT monitoring platform, including ServiceNow incident management and Dynatrace. In weeks three and four, AWS conducted AWS Well-Architected Framework reviews and operational readiness assessments across critical workloads running on Amazon Elastic Container Service (Amazon ECS), AWS Lambda, Amazon Aurora and Amazon DynamoDB. “We didn’t anticipate the depth of incident management integration or how closely the AWS team aligned with our workflows to ensure seamless data flow,” states Gabriel.

Following the readiness assessments, United Airlines conducted simulations and game days to test real-world incident scenarios and align people, processes, and tools. The exercises surface operational gaps ahead of go-live and validate end-to-end workflows under realistic operating conditions. Teams confirm the automated creation of critical incidents in ServiceNow, rapid alert ingestion, and seamless generation of AWS support cases. Araseli Cerda, director, Cloud and Platform Engineering at United Airlines, says, “When incident managers detect a P1 or P2 event, they can create an AWS support case directly within ServiceNow without switching systems or losing time.”

United Airlines reduced the operational impact of IT incidents across its global network. The airline improved mean time-to-engage (MTTE) for critical incidents by more than 70 percent and reduced mean time-to-resolve (MTTR) by 7 percent, helping teams respond faster during operational disruptions that could affect passengers and flight schedules. The addition of AWS Unified Operations, along with other internal operational improvements, contributed to these results, with AWS experts available within five minutes for high-severity events. Gabriel says, “AWS Unified Operations supports our teams so they can react more quickly during critical events, which directly supports the airline’s on-time performance and strengthens customer loyalty.”

By automating incident detection and escalation, United Airlines frees engineering capacity for higher-value innovation. AWS service impairment notifications flow directly into the airline’s monitoring platform, decreasing the need to investigate alerts in separate consoles. Incident managers enter response meetings with AWS runbooks preloaded with relevant business context, accelerating resolution and improving coordination. Cerda says, “We are resolving issues faster while strengthening our architecture to prevent future incidents.”

Furthermore, AWS Unified Operations supported United Airlines’ IT resiliency efforts while the airline continued to advance AI adoption on AWS. United Airlines automated more than 1,000 cross-regional failovers to enhance availability as part of its broader resiliency strategy. The airline also continued advancing AI adoption on AWS, using Amazon Bedrock, to support faster incident analysis and more efficient log interpretation during operational events. Gabriel says, “AWS helps us evaluate each manual step and determine whether it can be automated or enhanced with AI. These capabilities support us as we work to improve operational decision-making and work toward an even better passenger experience.”