Math & Statistics @ University of Toronto (ASIP Co-op) · data & quantitative analytics

I turn messy, unstructured data into decisions, building dashboards, data pipelines, and forecasting models. My background sits at the intersection of statistics and software engineering: I've shipped production systems professionally, and I apply that same engineering rigor to analytics work, from raw data all the way to stakeholder-facing insight.

Currently seeking a Fall 2026 data / quantitative analyst co-op.

• Analytics & BI: end-to-end pipelines from raw data to interactive Power BI dashboards
• Quantitative & statistical modeling: time-series forecasting, actuarial/mortality modeling, A/B testing
• Data engineering: SQL pipelines, ETL, and the software practices that keep analysis reproducible

Real Estate Market Intelligence Dashboard · Power BI SQL Python End-to-end analytics on 3,360 raw Calgary listings, engineered a 14-column semantic data model, authored 17 reusable DAX measures, and built a 5-page interactive report enabling self-service analysis across 309 neighbourhoods and 217 brokerages.

Mortality Forecasting for Insurance Risk · Python R PyTorch StMoMo Modeled the USA-Bulgaria life-expectancy gap using Human Mortality Database data. Benchmarked three approaches (a hybrid Lee-Carter + LSTM rotation, an LSTM baseline, and a GLM), cutting forecast error 64.6% vs. the GLM, with a classic Lee-Carter benchmark built in R (StMoMo) projecting life expectancy to 2050.

Patient Activity Analysis & Forecasting · SQL Python PostgreSQL Built an ETL pipeline ingesting 20+ patient time-series datasets into PostgreSQL, engineered features with window functions, and reduced forecast error (MAE) from a 68.5 baseline to 0.0085.

Real-Time Fraud Detection · Python FastAPI Redis A fraud-detection microservice with sub-100ms latency, using an Isolation Forest to flag anomalies in heavily imbalanced transaction data, with fail-open design for continuous availability.

Social Media Application · Java Led a six-person team to build a CLI social platform on Clean Architecture and SOLID principles; improved search accuracy from 49% to 81% with Levenshtein and Jaro-Winkler algorithms.

Analytics & BI  Power BI · Tableau · Excel · DAX · Microsoft Power Platform Languages & Databases  SQL (SQL Server · PostgreSQL · MySQL) · Python · R · Java · JavaScript · Scala Data & Statistics  Pandas · NumPy · scikit-learn · XGBoost · PySpark · time-series forecasting · statistical modeling Engineering & Tools  FastAPI · React · Node.js · Git · Docker · Azure · AWS

LinkedIn · rahdin.zaman@mail.utoronto.ca · Toronto, ON