Case 1 · Annual Telecom API Peak Load Analysis (Hadoop / Hive)
48-hour annual API peak-load analysis for a 42M-record/day telecom system. 15.3B yearly logs made Oracle infeasible — resolved via rapid Hadoop/Hive cluster cloning and parallel ETL.
Engineering & Data Projects — ETL, performance tuning, and monitoring at scale.
This section presents real-world engineering and analytics projects that I have designed, built, and maintained in production and production-like environments, spanning ETL pipelines, SQL performance optimization, Hadoop/Hive processing, Power BI DAX modelling, and large-scale data systems.
Each project reflects hands-on problem solving, system optimizations, and scalable designs across telecom and civic data domains.
48-hour annual API peak-load analysis for a 42M-record/day telecom system. 15.3B yearly logs made Oracle infeasible — resolved via rapid Hadoop/Hive cluster cloning and parallel ETL.
From monthly full-scan to daily partitioned checks; ~30× scan reduction.
Write-up in progress
Sub-second lookup without ES — index + SQL tuning.
Write-up in progress
120 hosts, shell orchestration, proactive incident catch.
Write-up in progress
6M users cutover — resilient ETL and data integrity checks.
Write-up in progress
Detect & auto-resolve issues before customer calls.
Write-up in progress
From T−2 to T−1, ensuring on-time billing.
Write-up in progress
Demographic and engagement analytics of youth services in Calgary — exploring equity and participation through data visualization.
Browse engineering projects alongside city dashboards in Civic & Social Analytics, Public Health Analytics, Crime & Disorder Insights, Housing Market Insights, and Economy & Energy. Read long-form analysis in the Blog or learn more about my background on the About page.