Migration et modernisation de la plateforme technologique boutique

Client

A boutique technology start-up revolutionizing the US cannabis sector with a customizable platform designed to empower multi-state operators with innovative tools for market success.

Business Challenge

• Application Complexity: Services like data hub, consumer app, sync engine, and recommender rely on disparate platforms (ECS, S3, Lambda, Firebase), creating integration challenges.
• Database Limitations: Aurora MySQL is costly, with security and design inefficiencies impacting performance and compliance.
• Authentication Control Issues: Lost admin access to Firebase authentication necessitates migration to a secure and manageable solution.
• Architectural Challenges: Inefficient design with public traffic communication, mixed environments, and S3-hosted web pages hinders scalability and security.

Approach

• Enhanced RDS Aurora MySQL with schema, index tuning, and version upgrades for performance and cost-efficiency.
• Data pipeline built on RDS Aurora MySQL for multi-table extraction and processing, automated by Glue triggers.
• AWS Glue jobs for data extraction, cleaning, and transformation with PySpark, storing processed data in S3.
• Glue crawlers automate schema discovery and updates in the Glue Data Catalog for efficient querying via Athena.
• Data encrypted in transit (SSL/TLS) and at rest (SSE-S3 or SSE-KMS), with IAM roles configured for least privilege access.
• CloudWatch logs capture Glue job execution, and security configurations comply with industry standards with periodic audits.

Transformational Effects

• 16% Faster Query Performance: Reduced average response times for critical workloads
• 25% Reduced CPU Usage: Optimized resource utilization during peak hours
• Enhanced Data Accessibility: Streamlined ETL pipeline for improved analytics and reporting.
• Cost Efficiency: Leveraged serverless technologies for better performance at lower costs.