Case Study

Few intriguing success stories from some great brands.

Uninterrupted live class streaming with 100k concurrent users. Delivered

About Customer

Physics Wallah is a multinational Ed-Tech unicorn headquartered in India, primarily focused on providing quality education content at an affordable price for JEE/NEET/NDA/GATE/UPSC aspirants. Currently, the company has a user base of over 6 million active students.

Problem Statement
  • data analytics

    Needed resource independence and isolation, the entire infrastructure (prod + non-prod environments) was held in a single AWS account.

  • quality

    Achieve uninterrupted live class streaming at a scale of 50K – 100K concurrent viewers.

  • Continuous security evaluation and traceability, Data Backups to ensure protection against unexpected events or losses.

  • Lack of alerting for incidents, warnings, cost alarms, operational processes etc.

Solution Offered!
  • Set up different AWS accounts with comprehensive segregation of controls. Implemented preventive/detective controls on every account with AWS Control Tower and the defined hierarchy of accounts with AWS Organization.
  • Setup AWS Guarduty and Security hub for centralized findings and reports of all accounts.
  • Created customized Grafana dashboards to visualize critical endpoint SLAs, application-specific metrics, and their dependencies.
  • Implemented comprehensive observability visualization and business intelligence monitoring dashboards with AWS Quicksight.
  • Implemented Cross-region, cross-account RDS backup using Terraform.
  • Integrated Kubecost into the Kubernetes environment to gain comprehensive insights into cost allocation and utilization of its components.
  • Deployed CID Dashboards in AWS to monitor cost utilization per AWS account, service, etc.

Final Outcomes
  • A unified framework for managing accounts at scale with centralized Access and Privilege management

  • Streamlined cost allocation, resource tracking, and security monitoring.

  • Extended resilience of the platform to be able to manage a scale of around 100K concurrent users.

  • Disaster recovery setup to handle partial or complete regional outages to prevent data loss.

Looking for a similar solution?

Know More