Incident investigation, root cause analysis, SQL/log forensics, and API troubleshooting for SaaS platforms — plus AI-assisted tooling to make support workflows faster and more consistent.
Signal detected — entering the investigation
01/Featured Investigation
A private prototype exploring how incoming Slack support conversations could be triaged, assigned, and answered faster — with AI drafting responses that a human engineer always reviews before they're sent.
Read the full case study →02/Workflow · Signal to Resolution
An incident has just arrived — signal captured and logged.
How big is the problem? Scope and business impact assessed.
An engineer's workstation — reproduce the failure and trace it across every layer.
Investigation path Report → Reproduce → Isolate service → Trace dependency
Correlate logs, queries, and API traces around the failing request.
User can't place a bet — isolate the underlying cause, step by step.
Prove the root cause rather than guess it.
Verification result Root cause proven, not assumed
Provider contacted → confirmed → resolved. Close the loop.
Communicate Provider resolves → Validate → Close
Root cause traced to provider-side rejection. Provider confirmed and resolved. Internal platform validated healthy throughout — recurring issue type, no new RCA required.
Timeline, root cause, and resolution logged for stakeholder visibility. Preventive action only tracked when the issue is new — this one is a known provider-side pattern.
03/Capabilities · Applied In Every Investigation
Incident investigation · root cause analysis · production issue monitoring · engineering escalation and coordination.
SQL / MySQL · GCP & log analysis · API troubleshooting · Postman · Firebase · database investigation.
Flow Log → Query → Correlate → Identify → Confirm
Support workflow automation · AI-assisted investigation and documentation · workflow prototyping.
Support question → AI drafts a response using shared context → an engineer reviews before it's sent.
An ongoing project — scans the official race-day website, compares it against the internal site, and flags anything support should know about before it becomes a ticket.
CHANGE DETECTED → SUPPORT TEAM NOTIFIED
04/Field Notes · Additional Case Files
A collection of write-ups walking through production-style incidents — symptom, investigation steps, root cause, and resolution.
Write-ups 02Practice case studies focused on diagnosing data and application issues through SQL — from query investigation to root cause in the data layer.
Practice 03A self-built dashboard for organizing SQL practice — queries, concepts, and problem sets — while building toward production-ready database investigation.
Learning tool 04A lightweight tool for tracking applications, follow-ups, and interview stages during an active job search — built to stay organized.
Personal tool