How GeoPITS Helped Cyces Cut Data Latency from 24 Hours to Under 1 Minute with a Real-Time GCP Pipeline
A seamless real-time analytics pipeline built using Pub/Sub, Dataflow, and BigQuery.
Overview
Cyces is an engineering-first product studio that helps companies build, enhance, and scale their technology with strong expertise in data engineering and cloud operations. Their ride-sharing platform in Southeast Asia handles thousands of real-time trip events every day, and as this data grew, they needed a modern streaming pipeline that could deliver instant insights, support accurate pricing, improve route decisions, and ensure a seamless experience for both riders and drivers.
The Challenge
Legacy batch ETL created 12–24 hour delays in access to analytics.
High operational overhead due to on-prem ETL servers.
Difficulty handling continuous streaming data from IoT devices.
Limited scalability and increased delays during peak traffic hours.
Our Solution
Built a real-time Pub/Sub → Dataflow → BigQuery streaming pipeline.
Implemented windowing and late data handling mechanisms for accuracy.
Integrated Cloud DLP to mask sensitive user and location data.
Enabled auto-scaling and secure access using IAM and VPC Service Controls.
Leveraged BigQuery optimizations - partitioning and materialized views - for faster analytics.
Benefits
Business Impact
Real-Time Insights
Reduced data latency from 24 hours to under 1 minute.
High Availability
Achieved 99.98% uptime with auto-scaling during peak usage.
Lower Costs
Cut operational overhead by 40% using serverless components.
Scalable Data Backbone
Built a streaming platform that easily handles millions of events per hour.
.svg){kind=icon}
