Real Projects
Every project here is a real system built from scratch — version-controlled, reproducible, and documented. Not demos. Not sandboxes.
Full migration of a 50-user organisation from Windows Server 2019 to Ubuntu 22.04 LTS — including Active Directory, email, file shares, and all business applications.
✓ Zero Downtime · Production LiveThe client was running expensive Windows Server licenses across 3 servers for file sharing, Active Directory, and print services. Licensing costs were unsustainable and the IT team had no in-house Linux expertise.
Complete AWS cloud architecture provisioned entirely with Terraform — HA, multi-AZ, least-privilege IAM, auto-scaling, and full observability built in from day one.
✓ Production Live · 99.9% UptimeThe client had manually configured AWS resources with no IaC, no consistent tagging, over-permissive IAM roles, and no disaster recovery plan. Infrastructure was undocumented and impossible to reproduce.
Multi-stage automated deployment pipeline that eliminated manual deployments entirely — from code commit to production in under 8 minutes with automatic rollback.
✓ Fully Automated · Zero Manual StepsThe development team was manually deploying via SSH — a 2–3 hour process prone to human error. Failed deployments required manual rollback. No staging environment existed. Production was always a risk.
Complete observability platform for a multi-server Linux environment — metrics, logs, dashboards, and alert routing all deployed from Ansible in under 30 minutes.
✓ Live Alerts · 24/7 MonitoringThe client had no visibility into their infrastructure. Outages were discovered by users, not by the engineering team. Logs were spread across multiple servers with no central aggregation or search capability.
Systematic CIS-aligned hardening of 8 production Linux servers using Ansible — automated, repeatable, and producing compliance reports for audit.
✓ CIS Compliant · Attack Surface −90%Default Ubuntu installations with root SSH enabled, no firewall rules, unnecessary services running, and no intrusion detection. Multiple servers — inconsistent configuration. One compromised server away from full breach.
Multi-container Docker system with automated failure detection, health checks, and recovery — 9 services running continuously with zero manual intervention.
✓ 99.8% Uptime · Zero InterventionClient ran 9 Docker containers with no health checks, no restart policies, and containers frequently crashing silently — requiring manual intervention to restart. Engineers were paged at night regularly.
Next Project
Every engagement starts with a free discovery call. No obligation — just a clear picture of what needs to be fixed and what it will take.