How SciPlay Saved Over $469K Annually on Real-Time Data Pipelines with Zipher

September 18, 2025

Company: SciPlay
Industry: Mobile Gaming
Headquarters: Las Vegas, Nevada, with US studios in Cedar Falls, Iowa and Austin, Texas.

Cloud: AWS
Challenge: Optimizing Databricks real-rime Data Pipelines for cost and reliability
Results: 35.7% cost reduction

The Challenge: Containing Real–Time Data Pipeline Costs in a Growing
Data Landscape

SciPlay, the social games division of Light & Wonder, develops and publishes popular free-to-play mobile games. The company leverages data at massive scale to gain user insights, power game personalization, and for real-time analytics. As SciPlay’s games’ adoption and monetization soared, their streaming workloads grew exponentially. This led to escalating Databricks and AWS costs, and a heightened risk of processing disruption and data latency.

Despite significant investment in Databricks on AWS, managing compute efficiency for hundreds of real-time data pipelines became increasingly complex and resource intensive. To address these challenges, SciPlay needed a solution that would reduce costs and improve reliability without disrupting their existing teams or data pipelines.

The Solution: Zipher’s Zero-Touch Databricks Optimization Solution

To meet this challenge, SciPlay deployed Zipher in their entire production environment, totaling hundreds of streaming jobs. Zipher’s automated optimization platform made it easy to get started, with a 5-minute install, zero code changes, and a secure setup that didn’t require access to SciPlay’s data.

Zipher applied advanced optimization features including:

  • Shifting from static to dynamic streaming infrastructure: Zipher applied dynamic cluster configurations that learn from individual stream patterns and automatically adapt to varying data loads.
  • Spark-aware autoscaling: Zipher’s proprietary autoscaling algorithm dynamically adjusts resource allocation at runtime, tuned specifically to the stream profile and needs.
  • Improved utilization of spot instances: Zipher improved efficiency through AI-powered availability-zone selection and real-time adjustments to spot capacity. 

“Zipher helped us cut costs dramatically while improving reliability on our streaming pipelines. The setup was painless, and the results were impressive.” – N. Amami, Director of Data Engineering, SciPlay

The Results: $469K Saved and Fewer Failures

Zipher’s impact was fast and measurable. After deploying across their entire environment,SciPlay saw significant improvements in both cost and performance:

  • 35.7% reduction in Databricks streaming job costs
  • 43.1% savings on AWS compute costs
  • 48% decrease in job failure rate
  • $469,000 in projected annual savings across optimized workloads

And all of this was achieved without changing a single line of code.

“With Zipher, we were able to achieve what every data leader wants – significant cost savings without compromising performance. It was like flipping a switch: no disruption, no code changes, just immediate efficiency. Zipher has become an important part of our data stack.” Y. Vidergor, VP Data, SciPlay

Conclusion: Game-Changing Cost Efficiency

For high-scale, real-time businesses like SciPlay, data infrastructure must be both powerful and cost-efficient. Zipher delivered both – enabling the team to optimize hundreds of jobs autonomously, cut compute waste, and improve streaming reliability.

Ready to see results like SciPlay’s? Book a time to discuss automated Databricks optimization