Master of Applied Computer Science @ Concordia. Data analyst and BI professional with hands-on experience in ETL pipeline design, star-schema modeling, DAX development, and executive dashboard delivery. Python | SQL | Power BI | Tableau | Azure
Professional Journey
Languages, Tools & Platforms
End-to-End Analytics Solutions
End-to-end data engineering pipeline built on the U.S. HMDA mortgage dataset (millions of loan records across 2013–2017). The goal was to ingest raw, multi-file government data, transform it through a medallion architecture on Azure Databricks, and surface business and fair-lending KPIs through SQL analytics including approval-rate gaps across demographic groups, the exact question the dataset was created to answer.
End-to-end cloud data engineering pipeline. Azure Data Lake Gen2 → Azure Data Factory (Spark) → 6-page Power BI dashboard. Ingests 4 formats (CSV, Parquet, Avro, JSON), ~180K records across 2015–2018. Azure SQL UAT validation. Delivery late rate: 54.52% flagged as critical insight.
246,741 chip transactions across 300+ stores. Customer segmentation across 21 demographic segments, statistical t-tests to validate price tiers (all p<0.05). Pearson correlation control-store matching (r=0.97) for trial store rollout decisions. 3-dashboard Power BI suite.
Montreal GTFS data. 11-table star-schema with custom engineered tables (fragility_index, service_week_profile). 3 Power BI dashboards with GIS/Esri map layer. 215 routes, 273,967 trips, 8,745 stops. Most complex data model in portfolio.
Full database management system in pure Python using CSV as storage engine. Implements ACID-like transactions (UUID staging dirs, atomic commit/rollback), SHA-256 auth, SQL-like query engine (SELECT / INSERT / UPDATE / DELETE / WHERE), and per-user directory isolation.
Python ETL on 55,500 patient records → 50,000 clean via 2-stage deduplication. Feature engineering: age bands, billing ranges, stay duration. 5-dashboard Tableau suite — clinical outcomes, financials, physician/hospital performance. $1.28B total billing analysed.
LendingClub 500K+ records. Chunked Python ingestion → 3-stage MySQL pipeline: risk state classification, composite early-warning signal engineering (high utilization, low buffer, high leverage), portfolio bucketing and exposure analysis by risk tier.
RBC banking dataset (10K customers). Full BI lifecycle: BRD authored → 5-dimension star schema → advanced DAX (MoM time intelligence) → 2-page executive Power BI dashboard → parallel Excel UAT validation with 100% metric match verification.
7 IMDb TSV datasets (10M+ rows). Rolling-window actor breakout detection, episode continuity anomaly detection, actor-director pairing via recursive CTEs for CSV-split directors, decade-level genre trends. All Python results cross-validated in MySQL.
PostgreSQL across 5 business domains: restaurant ops, banking, retail data mart, subscription analytics, pizza delivery. Techniques: CTEs, window functions (ROW_NUMBER, RANK, SUM OVER), cohort analysis, percentile calculations, conditional aggregations, LEAD/LAG.
Berka Czech banking dataset — 8-table relational schema (clients, accounts, loans, transactions, cards, districts). Full pipeline: schema design + LOAD DATA INFILE ingestion, gender/DOB decoding from encoded birth_number, 40+ SQL queries across 4 analytical domains: loan risk, transaction behavior, client segmentation, and district-level analysis. Fraud detection via velocity checks, dormant-active pattern flagging, round-number transaction detection, and 3× avg anomaly flagging. Advanced: 6-CTE Transaction Health Scorecard classifying accounts as Healthy / At-Risk / Declining using STDDEV, streak analysis, and LAG; full Customer 360 view combining age, gender, net balance, loan status, card type, and composite RANK score.
Data Analyst & BI professional open to opportunities in analytics, business intelligence, and data-driven roles. Experienced in ETL pipeline design, DAX development, and cross-functional stakeholder collaboration.