« CASE STUDIES

A data mesh platform

Building a data mesh platform with automated domain vending for a hyper scale gaming customer.

Tags

Data PlatformsData PlatformsModern Data StrategySnowflakeTerraformDevOpsGaming

Date

May 2024

Client

VGW

The Client

This innovative and highly profitable game design workshop, with tens of thousands of players and diverse product lines, stands out as a pioneer in social gaming.

The client’s platform is a highly scalable, highly available distributed system built on the AWS global cloud. With players all over the world, they leverage the power of AWS services to deliver a seamless customer experience via a rapid development lifecycle.

The volume of traffic and the demands of analytics, compliance and reporting on a gaming platform means that the systems generate millions of records every hour. This data originates from and is owned by various segments of the business, each presenting its own unique shapes and forms.

People having fun with social gaming

The Challenge

Building a domain driven scalable data platform with good security and RBAC structure was pivotal to the organisation’s strategy.

As the client embarked on their quest for consistency, improved security, and enhanced readability of their data, they sought specialised expertise to construct a secure, cohesive, and sustainable data platform. This platform would enable every business domain to integrate their data uniformly, leveraging the same standardised security posture across the board.

They turned to Mechanical Rock as a long-term trusted partner in software and data engineering to help them with this challenge.

The Solution

Mechanical Rock worked together with the client’s engineering team to build a cloud based secure data platform based on industry leading practices:

  • Continuous Integration and Delivery with security focused PR checks (CI/CD) of the entire stack including infrastructure and all data resources as well as automated non-functional tests to check RBAC and security measures on all PRs
  • Infrastructure-as-code deployment of all AWS and Snowflake resources, leading to more repeatable deployments, standardization and less configuration drift
  • Trunk based development for shorter lead times, faster feedback and more reliable change cycles
  • Sandbox to deploy all the changes into a sandbox environment in order to run all the checks and avoid production incidents
  • Strong security including integration with a third party identity provider and a new ‘least-privileged’ Roles Based Access Model (RBAC)
  • Self service provisioning of business domain resources to allow new business domains to on-board themselves into the platform via a predefined patterns
  • Consistent pre-defined access pattern for all business domains and all consumers across the board
  • Consistent pattern for data sharing across various business domains
  • Standardised tagging and masking policies to consistently and automatically apply data masking for all business domains based on a list of predefined tags
  • Consistent Dev, Test and Prod environments for all domains
  • Domain on-boarding quick-start documentation to allow speedy start
  • Automatic budget caping and alerting for all business domains
Domain on-boarding diagramDomain on-boarding diagram

The Benefits

  • Self service provisioning of business domains in less than 10 minutes
  • Improved security posture through strict access controls and automatic data masking
  • Consistent structure, access pattern and data sharing throughout the organisation
  • Averaged 4 production releases per day
  • Automated testing on every change to the system
  • Improved reliability through pipeline alerting & monitoring
Pipeline steps to deploy to prodPipeline steps to deploy to prod


THINK WE CAN HELP YOU?

Get in Touch

Reach out to us and a member of our team will be in touch right away.

contact@mechanicalrock.io