Modernizing Gathern on AWS and improving performance by 77%

When Gathern began operations in 2016, the company relied on a monolithic application built on cloud infrastructure. As the marketplace grew by over 500 percent in just 2 years, the architecture that once supported Gathern began to hold it back.

The tightly coupled system meant that scaling any single feature required scaling the entire application, increasing costs and slowing response times. Deployments became high-stakes events, which limited releases to just one or two per month. Meanwhile, guests wanted to search for accommodations using natural, conversational language in Arabic or English, but the existing filter-based system could not interpret these requests.

Gathern’s leadership recognized an opportunity to transform these constraints into competitive advantages. The company had built on AWS from nearly the beginning, drawn by its stability, security, and continuous innovation. By taking advantage of the broad portfolio of compute, database, and AI services on AWS, Gathern could unlock new levels of performance and improve the customer experience.

Gathern built a new microservices architecture in parallel with its existing monolithic system and gradually transitioned functionality to maintain zero downtime. The microservices architecture is based on Amazon Elastic Kubernetes Services (Amazon EKS) Auto Mode, which fully automates Kubernetes cluster management for compute, storage, and networking on AWS with a single click. Using this service, Gathern gained a resilient foundation for containerized workloads while simplifying cluster management and automatic scaling.

With the container orchestration layer established, the team built out supporting infrastructure. Amazon Simple Queue Service (Amazon SQS), a fully managed message queueing service, handles asynchronous communication between microservices, preventing cascading failures and smoothing traffic spikes. Amazon ElastiCache, a serverless, fully managed caching service, provides high-speed caching that reduces latency for frequently accessed data.

For the primary database, Gathern uses Amazon Aurora, which offers unparalleled high performance and availability at a global scale. Read/write traffic is split to prevent browsing activity from affecting booking performance. The team also migrated database and compute workloads to AWS Graviton processors, which are designed to deliver the best price performance for cloud workloads.

“In terms of performance and cost, we found a great improvement just by switching to AWS Graviton,” says Mohamed Mohsen, DevOps engineer at Gathern. “This was perfect.”

To enhance the guest experience, Gathern engaged Zero&One, an AWS Partner, to develop an AI-powered natural language search. AWS offers comprehensive generative AI capabilities, from high-performing infrastructure to streamlined application development with enterprise-grade security. Gathern used Amazon Bedrock, which powers generative AI for more than 100,000 organizations worldwide, to build a system that interprets guest queries in Arabic or English and converts them into structured search filters. The company also adopted Amazon Rekognition, an image recognition and video analysis service, to automate image moderation for host-uploaded photos.

By transitioning to a microservices architecture on Amazon EKS and migrating workloads to AWS Graviton processors, Gathern decreased its average API latency from 600 milliseconds to 140 milliseconds, a reduction of approximately 77 percent. In December 2025, the system handled peak traffic of 90,000 requests per minute without incident, demonstrating the resilience and elasticity Gathern needed to support its rapid growth.

The modernization has also unlocked a new pace of innovation. Previously limited to one or two major releases per month, Gathern can now deploy updates continuously. “Containerization on Amazon EKS has greatly enhanced our automation,” says Ehab Hakawati, director of engineering at Gathern. “We have been able to release 87 times in a single month.”

Gathern plans to complete its microservices migration in early 2026 and expand its use of generative AI, leaning on AWS to guide its adoption of new services and features. “I consider the AWS team to be more than just a provider,” says Hakawati. “They gave us a lot of ideas and insights, and we are capitalizing on them.”