Work

Problems I have owned end to end.

A few representative pieces of support and observability work — each a real situation, the action I took, and the result it produced.

Featured

2025

Problem: On-call needed to catch service degradation before customers felt it — with signals spread across three separate observability stacks.
Approach: Designed and maintained dashboards in Splunk (SPL), Dynatrace (DPL patterns) and Datadog (monitors and alerts), correlating application logs against service metrics in one view.
Outcome: Degradation was caught early, triage got faster, and escalations carried reproducible evidence instead of guesswork.

SplunkDynatraceDatadogMonitoring

Featured

2025

Problem: Recurring tickets were re-investigated from scratch, and newer engineers had no documented path through them.
Approach: Authored solution-based resolution and triage guides in the team knowledge base, turning hard-won incident knowledge into a repeatable procedure.
Outcome: New engineers ramped faster, time-to-triage on similar cases dropped, and senior engineers were freed for genuinely novel problems.

Knowledge BaseIncident ManagementMentoring

Featured

2025

Problem: Too many manual steps sat between an alert firing and a root cause being found, so investigations started from scratch.
Approach: Administered Linux hosts and wrote Bash scripts to automate health checks and log collection, alongside Docker-based services.
Outcome: Investigations began from collected evidence rather than a blank terminal, trimming the path from "alert fired" to "root cause found".

LinuxBashDockerAutomation

2024

Problem: Billing and reporting gaps on a 24/7 US healthcare SaaS traced back to third-party API integrations failing silently.
Approach: Analysed HTTP request/response flows and application logs to isolate the failing integrations and reproduce them on demand.
Outcome: Findings were packaged for engineering, and both technical contacts and non-technical stakeholders stayed informed through to resolution.

REST APIsLog analysisUS Healthcare

2024

Problem: Compliance reviews required tracing data discrepancies and proving record accuracy across multiple tables.
Approach: Ran multi-table JOINs, stored procedures and DML operations to follow the data, verify records and reconcile the gaps.
Outcome: Produced audit-ready reports that billing and clinical teams could rely on during compliance review.

SQLData IntegrityCompliance

Loading…