Chemical Engineer turned GA4 Data Integrity Auditor | Process Control, Mass Balance & First-Party SST

Most e‐commerce and SaaS companies lose 20–40% of their conversion data due to broken tracking and client‐side signal loss. But insights are only as good as the data infrastructure underneath them.

I don't approach tracking as a marketer; I approach it as a process engineer. I apply Mass Balance and Failure Mode Analysis (FMEA) to analytics infrastructure to find where the pipeline leaks and why — and now extend that into first‐party Server‐Side Tagging (SST) on GCP App Engine.

• GA4 Data Integrity Audits: 47‐point diagnostic across Shopify, GTM, and GA4 — scored matrix, dev‐ready fix tickets, and a QA monitoring dashboard
• GA4 ↔ Shopify Mass Balance Monitoring: FULL OUTER JOIN reconciliation between GA4 events and Shopify orders, with FMEA leak classification (FM‐01 tracker suppression, FM‐03 phantom purchases, FM‐05 sensor ghosts) and quantified impact
• First‐Party SST on GCP App Engine: Dual GTM Web + Server containers with custom first‐party endpoints (e.g., collect.aniji.ca) to bypass ITP/ad‐blockers and restore signal integrity, plus app.yaml + GTM JSON exports and a GitHub Pages trigger environment to validate Web → Server → GA4 delivery before production
• NL2SQL & Semantic Layer QA: Design of Experiments (DoE) stress‐testing of AI‐powered query platforms to diagnose architectural failures before they reach production
• SaaS Revenue Protection: Dual‐method SQL redundancy validation to identify at‐risk MRR before it disappears from your model

• GA4 ↔ Shopify Mass Balance Monitor & First‐Party SST Proxy — Diagnosed a 4.95% phantom ROAS over‐attribution rate, preventing an estimated ~80,000ドル in phantom revenue from contaminating media models. Includes a live Looker Studio monitor and full cloud config (SQL, FMEA, migration framework, GTM exports).
• GA4 Data Integrity Audit — 47‐Point Diagnostic + Dev‐Ready Fix Plan — Engineered a PIMS‐style validation pipeline with automated SQL event testing, delivering a 47‐point diagnostic matrix, prioritized fix tickets, and a QA monitoring dashboard.
• NL2SQL Semantic Layer Audit | 16 Failure Modes Diagnosed — Executed a 30‐query DoE across 5 bug classes on an enterprise NL2SQL platform. Real client, 5‐star review.
• SaaS Churn Risk Engine | ~250ドルK At‐Risk Revenue — Identified quarter‐million MRR exposure using engineering redundancy validation and dual‐method SQL cross‐checks. Random Forest ML + Tableau control panel.

GA4 Google Tag Manager Shopify BigQuery SQL Python Looker Studio Scikit-learn Tableau GCP App Engine Server-Side Tagging

• Upwork: Book a Consultation / Audit
• LinkedIn: Charles Aniji