Customer Story
How a global cruise operator rebuilt its digital storefront on Akka — releasing 4× more often and running for years without a single crash.
Norwegian Cruise Line ran on three backend systems and five codebases in multiple languages, deploying at midnight to avoid disruption. Rebuilt on Akka, the company now releases 4× more often — completing 65+ deployments in 12 months with zero rollbacks, running shipboard payments for over two years without a single crash, and deploying during business hours as it posted record profits.
Akka ROI Scorecard
more frequent releases. Deployments moved from weekly and monthly cycles to a continuous cadence — 65+ shipped in 12 months with zero rollbacks.
lower run cost. Three backend systems and five codebases in multiple languages consolidated onto one platform, cutting duplicated infrastructure and maintenance.
the wave-season booking peak absorbed on the primary platform, with shipboard payments running 2+ years without a single crash.
Business outcome: record profits, with midnight deployment windows retired for business-hours releases.
Norwegian Cruise Line ran its digital business on three backend systems and five codebases written in multiple languages, wired together as a monolith with complex dependencies. The architecture slowed the developer team and stretched release cycles to weekly or monthly. Every deployment had to run at midnight to keep the risk of disruption away from customers.
Demand is highly seasonal. Wave season — the early-year window when a large share of annual bookings land — drives sharp traffic spikes that the legacy platform struggled to absorb without added cost and risk. Norwegian set two goals: more agility for the developer team, and more frequent release cycles to support the business.
Norwegian rebuilt its codebases and converted the monolith into a service-oriented architecture with decoupled systems on Akka. The majority of ncl.com and the partner API front-end now run on it. The deciding factors were agility, decoupled deployment, and resilience under seasonal load.
The Never Fail guarantee means Akka handles clustering, resilience, durable in-memory state, and traffic steering, so the application does not have to. Norwegian ran shipboard payments for the casino, arcade, and vending machines on it for over two years without a single crash.
Scale. Decoupled services let Norwegian absorb wave-season booking peaks on the primary platform instead of provisioning around them. Shipboard payment systems for the casino, arcade, and vending machines ran 2+ years without a single crash.
Cost to operate. Consolidating three backend systems and five codebases in multiple languages onto one platform cut duplicated infrastructure and maintenance overhead by an estimated 45%, with fewer moving parts to run.
Speed to production. Release frequency rose 4×, from weekly and monthly cycles to a continuous cadence. Norwegian shipped 65+ deployments in 12 months with zero rollbacks and retired midnight deployment windows for business-hours releases.
Business impact. Norwegian Cruise Line posted record profits as its digital storefront moved faster and stayed available through peak demand.
Norwegian's booking surfaces — ncl.com and the partner API front-end — are exactly where agentic travel now lives: conversational and API-driven channels where an AI agent takes the booking, prices cabins dynamically against wave-season demand, upsells excursions and stateroom upgrades, and rebooks itinerary changes in real time. On the Akka Agentic AI Platform, those channels become governed AI agents running on the same decoupled, resilient services Norwegian already proved — with durable state that survives failure, sub-minute recovery, and no midnight maintenance windows.
The platform that rebuilt Norwegian's storefront is the same platform that builds and runs agentic AI today. Enterprises build on the Akka Agentic AI Platform directly, or have a system delivered and operated through Akka Specify. Each path provides the production guarantees Norwegian has run on for years.
