Parsa Hassas

About Me

Structured thinking, practical analytics

I'm a Business Analytics student at Seneca Polytechnic with a B.Eng. in Electrical & Electronics Engineering from Iran University of Science & Technology (IUST). My engineering background shaped how I approach problems: systematically, with attention to data quality, model validity, and communicable results. I'm currently focused on analytics projects involving data preparation, regime-based modeling, machine learning, and dashboard design — and I'm actively looking for a role where I can apply these skills in a real-world data team.

Skills & Tools

Technical Stack

Languages, frameworks, and platforms I work with.

Languages

Python SQL SAS

Data & ML Libraries

pandas NumPy scikit-learn NLP / Embeddings Matplotlib / Seaborn

Statistical Modeling

OLS Regression Ridge & Lasso Stepwise Selection K-Means Clustering PCA

Visualization & BI

Tableau Power BI Dash / Plotly Excel & Cognos

Data Engineering

API Integration Data Pipelines ETL & Cleaning Feature Engineering Databases (SQL)

Tools & Workflow

Git & GitHub Jupyter VS Code Microsoft Excel Agile

Featured Work

Projects

End-to-end analytics projects — from raw data pipelines to deployed tools and actionable insights.

PROJECT 01

TTC Ridership Recovery & Changing Travel Patterns

Regime-Based Transit Analysis · Python · SAS · Tableau · Dash

A structured analytical study of Toronto TTC monthly ridership across three COVID-era regimes. The project goes beyond descriptive charts to reveal how the structure of demand fundamentally shifted — which variables drove ridership, and by how much, changed entirely depending on the period.

Built a 179-observation monthly dataset by merging TTC ridership, gas prices, unemployment rates, and weather data from four open government sources.

Applied OLS regression, stepwise selection, Ridge, and Lasso independently across three regime subsets (Pre-COVID, COVID/Recovery, Post-COVID) to compare variable importance.

Post-COVID ridership recovered to 76.5% of pre-COVID levels — a structural shift, not just a volume gap.

Designed two Tableau dashboards (executive and technical) and deployed an interactive Dash app on Render for live exploration.

Average Monthly Ridership by Regime

Key finding: Pre-COVID ridership was governed by annual seasonal memory (ridership_lag12). During COVID/Recovery it became reactive to gas prices and short-run momentum. Post-COVID shows partial normalization — seasonal structure returned, but with new unemployment and weather sensitivity. The pandemic didn't just reduce ridership; it changed what drives it.

Python SAS (PROC REG) Tableau Dash / Plotly Ridge & Lasso Feature Engineering Regime Analysis Time Series

GitHub Repo Live Dashboard

Transit Analytics

PROJECT 02

The Watchlist — Movie Recommendation System

Personalized ML Recommender · Python · NLP · Clustering · API Engineering

An end-to-end personalized movie recommendation system built from scratch to address the limitations of platform-specific recommenders: limited cross-platform discovery, filter-bubble behavior, and poor explainability. The system models user taste across multiple cinematic dimensions and ranks candidates from a 50,000-title pool.

Engineered an automated multi-source data pipeline integrating TMDb API, Wikipedia plot summaries (with multilingual fallback & translation), and IMDb ratings across a 50,000-title candidate pool.

Built a scalable two-pass enrichment architecture: fast TMDb enrichment for all 50K rows, then selective Wikipedia enrichment only for low-confidence records — minimizing API cost while maximizing coverage.

Created four aspect-specific text representations per movie (theme, mood, style, context) and computed dense semantic embeddings for each channel separately.

Modeled user taste via centered rating vectors and soft overlapping "taste streams" using PCA + K-Means clustering — capturing that users have multiple preference lanes, not one flat profile.

Final scoring framework combines aspect-level similarity, stream alignment, negative-preference penalties, and IMDb-based cinematic quality weighting for selective, explainable recommendations.

Key outcome: The system successfully moves beyond single-platform, trend-driven suggestions by modeling taste as overlapping preference streams rather than a flat profile. Recommendations scored against both personal fit and cinematic quality — surfacing niche films a standard recommender would never surface, with explainability built into every ranking decision.

Python NLP / Embeddings K-Means Clustering PCA API Integration Data Pipeline Recommender Systems TMDb / IMDb / Wikipedia

GitHub — Coming Soon

Machine Learning

Work History

Experience

Professional background spanning engineering, data analysis, and editorial leadership.

Mar 2025 – Jan 2026 · 11 months

Customer Service Representative

TAGHZIE Electronics · Canada

Evaluated and analyzed client profiles to identify potential risks and ensure accurate data reporting. Researched market trends and conditions to inform client interactions. Maintained detailed documentation of customer inquiries and resolutions, ensuring organized workflows.

Jun 2023 – Dec 2024 · 1 yr 7 mo

Electrical Engineer

Power Supply Production · Tehran, Iran

Designed and tested electrical circuits, ensuring compliance with industry standards and safety regulations. Analyzed performance metrics and optimized product designs to enhance functionality. Collaborated with cross-functional teams to streamline production processes and reduce costs.

May 2023 – Dec 2024 · 1 yr 8 mo

Editor in Chief

Sarv Student Magazine — Culture, Art & Literature

Analyzed and reviewed data for discrepancies in editorial workflows, ensuring accuracy and integrity. Created documentation to streamline internal processes and enhance operational efficiency. Conducted detailed research on cultural and market trends to ensure high-quality publications.

Academic Background

Education

Interdisciplinary foundation combining engineering rigor with applied business analytics.

Postgraduate Certificate

Seneca Polytechnic

Business Analytics

Started January 2026 · In Progress · Toronto, Canada

Bachelor of Engineering

Iran University of Science & Technology (IUST)

Electrical & Electronics Engineering

September 2019 – January 2024 · Tehran, Iran