Understand an unfamiliar system
Map what is running, how services depend on each other, and what changed—even when documentation is missing or out of date. Give your team a shared picture of the system and where to investigate.
AWS · Platform · Networking
Production troubleshooting.
I help founders and engineering teams investigate slow, unstable or undocumented systems. I connect evidence across infrastructure, databases and applications, work with your engineers to resolve the cause, and leave the system easier to understand and change.
AWS CertifiedDevOps Engineer – Professional
AWS CertifiedSysOps Administrator – Associate
Where I can help.
Visibility across the stack
Connect metrics, logs and traces across infrastructure, databases and application requests. Give engineers the evidence to distinguish traffic growth from a slow query, leaking memory or a CPU-heavy code path.
Application & database troubleshooting
Investigate recurring outages, slow endpoints, memory leaks, inefficient queries and database contention. Work with your developers to isolate the cause and check that the change addresses it.
Infrastructure & network troubleshooting
Trace failures across compute, containers, DNS, routing and dependencies. Separate capacity and configuration problems from application behaviour before adding resources or changing the architecture.
Reviewable infrastructure changes
Bring existing AWS resources under Terraform and version control in stages. Make configuration changes and drift visible, with reviewable plans and checks before touching live resources.
Technical experience behind the work
OpenTelemetry, CloudWatch, AWS X-Ray, Grafana, RDS Performance Insights, PostgreSQL telemetry, Linux, Docker, ECS, EC2 and Terraform.
Instrumentation follows the investigation. I use or extend the tools your team already has.
Results from client work.
High CPU. Normal traffic.
I compared request volume with application telemetry and narrowed the CPU pressure to a specific endpoint.
Cause identifiedA CPU-consuming software loop.
The developers had a concrete cause to address.
Database performance
35–40%lower database load
I worked with developers to isolate inefficient queries and application behaviour using database telemetry, reducing memory pressure, deadlocks and customer-facing timeouts.
Recurring production outages
Memory leak identifiedthrough distributed tracing
I built end-to-end observability and used the traces to isolate a long-standing application memory leak behind recurring outages.
How I deliver.
Establish the facts
Map the system, agree the impact and gather a baseline. Add the visibility needed to investigate.
You getA system map & baselineNarrow down the cause
Test explanations against metrics, logs and traces with your engineers, separating symptoms from the cause.
You getEvidence & recommended changesResolve & verify
Make reviewed changes with a rollback plan, alongside your engineers. Check the outcome against the baseline and document what changed.
You getValidated changes & handover
Let’s talk
What’s happening in production?
Tell me what is slow, unreliable or difficult to understand, and what you’ve already tried. We can discuss whether I can help.
Your work isn’t passed down to another subcontractor.
I’m available for short projects, fixed-scope work and longer-term engagements.
I’ll reply personally.