Silicon Overdrive architected and built a new AWS best-practice environment, covering both Staging and Production and migrated Oye’s application onto it. Key elements of the solution included:
Compute: Amazon ECS on AWS Fargate was used for serverless, scalable container orchestration, removing the operational overhead of managing servers directly.
Database: Amazon RDS for PostgreSQL (engine 17.x) was deployed as the primary database, with native PostgreSQL replication used to migrate data safely from Azure into AWS.
CI/CD: GitHub Actions workflows were extended with OIDC federation, giving each ECS service its own scoped IAM role for short-lived, keyless deployment authentication.
Infrastructure as Code: Terraform was used to build the environment as a redeployable package across reusable modules, delivered to Oye as a project artifact.
Networking and delivery: Amazon Route 53 handled domains, Amazon CloudFront handled content delivery, and AWS Certificate Manager issued SSL/TLS certificates, with HTTPS enforced on both the ALB and CloudFront.
Container registry: Amazon ECR was configured with per-service repositories, access control, and image tagging.
Security: AWS WAF was deployed with six AWS-managed rule groups, including the Common Rule Set, Known Bad Inputs, Amazon IP Reputation List, Linux, Unix, and SQL injection rule sets, protecting both CloudFront and the public ALB. Amazon GuardDuty was enabled across S3, EBS malware protection, and ECS Fargate runtime monitoring. AWS Config was enabled with conformance packs covering IAM, S3, and security-services best practices, validated using Trend Vision One™ Container Security and Cyber/Cloud Risk Exposure Management (CREM).
Resilience: AWS Backup was implemented with daily, weekly, and monthly backup plans, a dedicated backup vault, and 30-day RDS backup retention. Database restore was carried out by Oye’s own technical team using an external tool, with Silicon Overdrive providing advisory support.
Because the entire environment had been built in Terraform as reusable, redeployable modules, Silicon Overdrive was able to stand up equivalent Production and Staging environments in the AWS Europe (Ireland) Region ahead of the planned cutover, directly in response to the AWS-reported regional disruption in Bahrain with no impact to Oye’s migration timeline or live services. The original Bahrain-region resources could not be decommissioned via Terraform at the time, since the same regional outage that prompted the move also affected control-plane access to that region; infrastructure teardown was deferred to Oye’s team until AWS access to the region was restored, while cutover to the new environment proceeded independently and without delay.
Following the initial migration, Oye requested a USSD access channel for Songa Na Oye. Silicon Overdrive integrated a new USSD-API ECS service with Oye’s existing ECS APIs, including a dedicated ECR repository, IAM roles and policies, a private ECS service, and updated Application Load Balancer routing to allow Oye to route traffic to the USSD-API container.