About ThetaRay

 We’re transforming the way that the world benefits from data. ThetaRay’s AI solutions pinpoint future problems, discover new opportunities, and expose blind spots within complex data sets that are growing bigger all the time. Our patented algorithms detect anomalies in real-time, radically reduce false positives, and uncover “unknown unknowns”

Industry: Financial Services

Primary project location: Israel

About Sela Software Labs Ltd

More (by Sela group) is an international team of Multi-Cloud Experts, delivering the best Multi-Cloud services to its customers and helping them grow.

Start Date: 1.09.2020 | End Date: 30.09.2020

 Partner Role in Project

AS partners we analyzed the current architecture and challenges faced by Thetaray. We then came up with the solution for containerization and automated deployments to address cost challenges. We then helped create a plan for implementation and provided assistance during the implementation and testing phase.

The challenge

Thetaray was facing a high cost of running their applications on virtual instances. To make sure the infrastructure was used in the most cost effective way was the primary challenge, even if it meant modernizing the application and the infrastructure.The production workload would continue to run on-premise due for regulatory reasons, however dev/test environments and a setup to run difference POCs in a cost-effective manner was needed.

The solution

The recommendation for Thetaray was to modernize their application and infrastructure. The use of GKE along with preemptible VMs helped the application scale when required while also keeping the costs low. Since the production would remain on-premise, deployment automation was necessary. This was achieved using Jenkins deployed in a separate Kubernetes cluster.

Google Cloud Features Used:

Kubernetes EngineCompute Engine

VPC Peering 

The results

  • Cost Reduction of 25% – Using GKE and preemptible VMs for dev/QA and build environments helped bring down the cost by 25%.
  • Streamlined infrastructure management – AS part of automated deployment process a labeling mechanism was added. This created labels for infrastructure that helped identify resources that could be deleted based on idle time and test cycles.
  • Modernization and Efficient Scaling: Containerization and use of GKE ultimately lead to the application running in a very efficient manner.

Customer Feedback

With our infrastructure growing significantly, we needed a partner who can support us and give us fast and professional results. With MORE and google cloud we can be sure that our big data analytics platform will run smoothly and securely.

Asaf Zinger, SVP R&D, Thetaray