Case study · Events · AWS
Memorizz Cut Deployment Time by Over 60% with AWS CloudFormation
Memorizz is a web and mobile application for creating, managing and sharing personalized events, media files and photobooks. As it scaled and added AI-powered captioning with Amazon Bedrock and AWS Lambda-based image processing, its manual infrastructure setup caused problems.
- Over 60% less deployment time
- One stack each for Beta, Stage and Production
- Stacks redeployed within minutes

The Challenge
Memorizz needed automated, repeatable and auditable infrastructure provisioning, so it could scale reliably while maintaining performance, compliance and cost-efficiency.
Inconsistent Environments
Keeping Dev, Stage and Prod environments consistent, with uniform configurations, was difficult.
Error-Prone Provisioning
Manual provisioning was error-prone and caused discrepancies between deployments.
Slow Updates and Scaling
Infrastructure updates and scaling tasks were time-consuming.
No Version Control
Infrastructure changes had limited visibility and no version control.
The Solution
HabileLabs moved Memorizz's AWS infrastructure deployment process to AWS CloudFormation, following Infrastructure as Code (IaC) best practices. The goal was to standardize deployments across environments, remove manual intervention, and enforce security and scalability through CloudFormation templates and AWS-native automation tools.
Infrastructure Automation with CloudFormation
All core infrastructure was defined in modular CloudFormation templates:
- VPCs, subnets, route tables and internet gateways
- Amazon ECS clusters and services on AWS Fargate for Beta, Stage and Production
- Application Load Balancer (ALB), target groups and listeners
- Amazon S3 buckets for original and resized images
- Amazon DynamoDB tables and Amazon RDS for MySQL configuration
- Amazon CloudWatch logs, metrics and alarms for monitoring
- IAM roles, policies and secrets management
CI/CD Integration
A deployment pipeline built on AWS CodePipeline and AWS CodeBuild, itself defined in CloudFormation, builds container images from the source repository (Bitbucket) and deploys them to Amazon ECS.
- Each stage (Source → Build → Deploy) was codified for transparency and auditability
Multi-Environment Support
Each environment is its own stack:
- Beta, Stage and Production each got a separate CloudFormation stack
- Parameterized templates hold environment-specific settings such as CPU and memory, subnet mappings, image tags and scaling policies
Monitoring, Logging and Alerting
CloudFormation templates provisioned:
- CloudWatch alarms for Amazon ECS, AWS Lambda, Amazon API Gateway and Amazon RDS metrics
- Log groups for Amazon ECS and AWS Lambda to centralize logs
- AWS X-Ray tracing to monitor user flows
Security and Compliance Automation
Security is provisioned with the rest of the stack:
- IAM roles with least privilege, AWS Secrets Manager entries and AWS KMS encryption keys, provisioned automatically through CloudFormation
- Security groups and NACLs that enforce strict inbound and outbound rules in line with NETSEC-001 guidelines
Business Impact
Operational Efficiency
Standardized templates reduced deployment time by over 60% across environments.
Consistency and Repeatability
Each environment mirrors the others exactly, reducing bugs and deployment mismatches.
Rapid Recovery
If something fails, entire stacks can be redeployed from CloudFormation templates within minutes.
Scalability
Auto Scaling policies and containerized Amazon ECS workloads scale with load.
Auditability and Governance
Every infrastructure change is tracked and version-controlled through Git-based CloudFormation templates.
Conclusion
Planning something similar?
Using AWS CloudFormation and other AWS-native DevOps tools, HabileLabs delivered a fully automated, secure infrastructure deployment for Memorizz that scales with load. The work also aligned Memorizz with AWS Well-Architected best practices and CloudFormation Service Delivery expectations.