I measure the existing system before rebuilding it
Slow performance, rising cloud bills or outages may have more than one cause. I review application flow, infrastructure, traffic, data, releases and current measurements together. Where I have access, I test hypotheses against data; otherwise I make assumptions explicit.
The aim is not simply to make one screen look faster. I balance user experience, operational reliability, security, maintenance and cost.
Performance, traffic & reliability
High traffic does not call for the same answer in every system. First find where requests wait: application code, database, external services or the release chain. Caching, query design, CDN/edge, background jobs and architectural separation should be justified by measurement.
- Baseline response time and error rate measurements
- Critical user journeys and dependency mapping
- Monitoring, alerts, incident management and rollback
- Maintenance, versioning and capacity planning risks
Cloud cost: visibility before optimisation
I examine AWS and GCN usage by service, environment and workload to understand where costs come from. Idle resources, poor sizing, unnecessary transfer or storage, licences and observability can all be worth reviewing. Real opportunities only emerge from billing and usage data.
I weigh each recommendation against performance, resilience, security and maintenance as well as cost. I do not promise a fixed saving; the baseline and success measures depend on the scope.
Possible deliverables
- A summary of current infrastructure and traffic flows
- A measured performance and cost baseline
- Risk, dependency and security checklist
- A prioritised improvement plan with impact and trade-off notes
- Recommendations for monitoring, releases, rollback and operations