What I do
I troubleshoot networking problems across AWS-hosted infrastructure, on-premises Cisco builds, and network/security appliances — Palo Alto firewalls, Nginx, Linux servers. I write Python for AWS Lambda automation and Bash for Linux compliance checks, and I use CloudWatch, Datadog APM, VPC Flow Logs, and Wireshark packet analysis to drive root cause. I've authored runbooks and methods of procedure, mentored junior engineers, served as an escalation point on 24/7 on-call rotations, and delivered executive-ready RCA summaries during major incidents.
Technical skills
Networking — protocol level
Network build & appliances
Automation & scripting
Cloud
Observability & RCA
Process, data & security
Professional experience
- Advised enterprise customers on cloud architecture, networking design, and migration decisions — resolving 200+ VPC routing issues and 300+ ALB/NLB/GWLB cases, improving load-balancer experience by 96%.
- Diagnosed Linux memory leaks, I/O latency, and storage failures using systemd logs, iostat, dmesg, and kernel metrics, cutting recurring production issues by 30%.
- Automated repetitive support tasks using Python and Bash, reducing manual operational overhead by 40%.
Selected projects
Serverless AWS cost optimisation pipeline
Built a serverless pipeline using Python, AWS Lambda, Cost Explorer, CloudWatch APIs, EventBridge, and API Gateway to detect idle EC2 resources and surface actionable savings recommendations.
CloudWatch Agent deployment & automated service-failure detection
Deployed and configured CloudWatch Agent to stream systemd, application, and system logs, resolving fragmented Linux telemetry. Built Python Lambda automation to detect systemd failures and I/O errors, with custom metrics and alarms for early anomaly detection.
Distributed observability dashboard (Datadog + CloudWatch)
Consolidated Datadog APM, CloudWatch metrics/logs, and Linux signals into a single view — integrating ALB logs, VPC Flow Logs, Lambda logs, and EC2 metrics via Kinesis Firehose pipelines. Built monitors for 5xx errors, packet drops, I/O spikes, GC pauses, and container OOM events.