City traffic at night

Case Study

Grab uses AutoMQ to replace Apache Kafka

~50%

Cost reduction compared to traditional Kafka

10x

Faster partition reassignment operations

Zero

Manual intervention for scaling operations

David Chen
"AutoMQ has been a game-changer for our streaming infrastructure. The operational overhead we used to spend on managing Kafka clusters is now completely eliminated. Our teams can focus on innovation instead of firefighting."

David Chen

Senior Engineering Manager, Platform Infrastructure

Grab

The Challenge

As one of Southeast Asia's leading super-app platforms, Grab processes billions of events daily across ride-hailing, food delivery, and financial services. Their Apache Kafka infrastructure faced significant operational challenges:

  • High operational overhead managing dozens of Kafka clusters across multiple regions
  • Expensive disk-based storage requiring constant capacity planning and over-provisioning
  • Slow partition reassignment operations causing service disruptions during scaling
  • 24/7 on-call requirements for manual intervention during traffic spikes
  • Increasing infrastructure costs as data volumes grew exponentially

Why AutoMQ

Grab chose AutoMQ for its cloud-native architecture and operational simplicity:

  • Serverless Architecture: AutoMQ's diskless design with S3 storage eliminated capacity planning and reduced costs by approximately 50%
  • Zero Operations: Automated partition management and self-healing capabilities freed engineering teams from on-call burdens
  • Instant Scaling: AutoMQ's architecture enabled 10x faster partition reassignment, supporting Grab's dynamic traffic patterns
  • 100% Kafka Compatible: Seamless migration without application code changes, reducing implementation risk
  • Multi-Cloud Ready: Native support for AWS and GCP aligned with Grab's multi-cloud strategy

The AutoMQ team worked closely with Grab's engineering teams to ensure a smooth migration, providing dedicated support and custom optimizations for their specific workload patterns.

The Results

Since migrating to AutoMQ, Grab has achieved remarkable improvements:

Key Metrics

~50%

Cost reduction compared to traditional Kafka

10x

Faster partition reassignment operations

Zero

Manual intervention for scaling

100%

Kafka API compatibility

Beyond the numbers, AutoMQ has fundamentally transformed how Grab operates its streaming infrastructure:

  • Engineering teams redirected from operational tasks to building customer-facing features
  • Eliminated on-call rotations for Kafka operations, improving team work-life balance
  • Seamless handling of traffic spikes during peak hours without manual intervention
  • Simplified multi-region deployments with consistent behavior across AWS and GCP
  • Improved system reliability with AutoMQ's self-healing capabilities

Ready to transform your streaming infrastructure?

See how AutoMQ can help you achieve similar results. Get a personalized demo and pricing comparison.