Systems we were trusted with, in our clients’ words.

My Interview Practice · multi-year engagement

A monolith broken up, and a platform that could finally be watched.

The platform began on a single outdated dedicated server — one machine to lose, manual deploys, and a security posture that had accumulated rather than been designed.

We migrated it to AWS, then led the transformation of the monolithic application into a microservices-oriented architecture built to scale for thousands of concurrent realtime users. Alongside it we built centralized monitoring, metrics and alerting, so production had visibility, incidents had a response path, and reliability became something measurable instead of hoped for.

• Monolith decomposed into a microservices-oriented architecture
• Designed to scale for thousands of concurrent realtime users
• Centralized monitoring, metrics and alerting for production visibility
• Migration from a legacy dedicated server to AWS
• EKS with horizontal pod autoscaling tuned to real traffic
• OpenTelemetry instrumentation with distributed tracing across services
• CI/CD in GitHub Actions with Snyk scanning in the delivery path
• Cloudflare hardening: DDoS and SYN/ACK mitigation, API rate limiting

Encirca · two-year engagement

Off shared cPanel hosting, onto Kubernetes — while holding SOC 2.

Encirca ran on cPanel web hosting: the arrangement that works until the business outgrows it, at which point scaling means buying a bigger box and every deployment is performed by hand. The move to Kubernetes had to happen without interrupting a live service, and without disturbing a compliance posture the business depended on.

We rebuilt it on Amazon EKS, then built the observability and security layers that had never existed — work that fed directly into the organization’s successful SOC 2 certification renewal.

• Infrastructure as code in Terraform, validated by Terratest before changes reached anything real
• CI/CD from commit through Helm, with canary deployments exposing releases gradually
• Observability built from nothing: Prometheus, Alertmanager, Loki and Grafana
• Real-time SLI/SLO dashboards with incident routing into Slack
• Automated security checks via AWS Security Hub, Inspector and GuardDuty
• Single sign-on consolidated across Auth0, Jenkins, AWS IAM with SAML, and WHMCS
• Multi-AZ RDS with automated backups for failover and recovery

Al Nassr Football Club · cloud engineering

Ticketing infrastructure for a club whose traffic arrives all at once.

Online ticket booking is one of the least forgiving workloads there is. Demand is not steady — it arrives in a spike the moment a match goes on sale, the transaction is financial, and the audience is public and vocal when it fails.

We supported the club’s online ticket-booking systems as cloud engineers: designing scalable infrastructure patterns for the load profile, and building the monitoring and operational visibility needed to hold availability and performance through the peaks.

• Containerized workloads on Docker, orchestrated with Kubernetes
• CloudFront CDN with geographic content localization to cut latency
• AWS WAF blocking OWASP Top 10 threats, with rate-based DDoS rules
• Datadog across the estate for log analysis and anomaly detection
• Ansible playbooks automating OS patching and access control
• On-call rotation with documented incident response in Confluence
• Cost analysis and budget alerting; RDS Reserved Instance optimization

WALT Labs · production support

Production support across a mixed estate of client environments.

Not every engagement is a build. WALT Labs needed dependable production support across a varied set of client environments, where the work changes shape daily and the environments are never uniform.

We covered CMS support, cloud operations, monitoring, troubleshooting, web functionality support and publishing assistance — including client-facing technical execution, where the person doing the work also has to explain it to the customer.