Toronto, Canada

Hi, I'm Ermias Gelaye Gaga

Data Scientist & Research Analyst

10+ years in research and analytics, from public-opinion monitoring to collective-bargaining analysis and AI-assisted document extraction.

  • M.S. Data Science
  • M.S. Cognitive Science
  • Senior Analyst
Portrait of Ermias Gelaye Gaga 10+ years in data

Professional path

About me

I'm a data scientist and research analyst with 10+ years of experience collecting, cleaning and analyzing data in fast-paced digital environments.

I hold a Master of Science in Data Science from Eastern University, a Master of Science in Cognitive Science (Cognitive Neuroscience stream) from CIMeC – Center for Mind/Brain Sciences, University of Trento, and a Certificate in Data Analytics and Visualization from the University of Toronto School of Continuing Studies.

I bring together psychology, cognitive neuroscience and data science to help research teams turn heterogeneous data into clear, defensible insights. Selected projects are below, each with a live demo and its source code.

Beyond data

Away from data, I photograph landscapes and travel across Canada and beyond.

10+Years of experience
3Degrees
20+Projects on GitHub
950+GitHub commits

Core competencies

Technical skills

What I work with, grouped by how I use it, with the projects that show it.

R

Also experienced

Statistical analysis and reporting with ggplot2.

Statistics ggplot2 Reports

Java

Also experienced

Backend systems and data processing.

Backend Systems Processing

Modeling & machine learning

Classification, cross-validation and unsupervised structure.

Scikit-learn Logistic regression Cross-validation PCA k-means

See it in Mental Health Google Health PCA

Tools used in these projects

Pandas NumPy Scikit-learn TensorFlow Matplotlib Seaborn Plotly D3.js Tableau ggplot2 PostgreSQL Flask REST APIs Git & GitHub Jupyter

Data science portfolio

Featured projects

Each data project has a live demo and its source code.

Map dashboard of Google health search trends across US statesData analysis · interactive map

Google Health Search Trends

Interactive dashboard development

Where Americans search Google for nine health conditions across 210 metro areas (2004–2017), which health worries travel together, and how CDC death rates changed over the same period.

  • Modeled Google Trends and CDC data in PostgreSQL, then rebuilt it as a serverless static site
  • Audited the data: fixed 18 wrong state assignments, 23 mis-geocoded metros and a CDC typo
  • Compared search interest with CDC BRFSS survey data: state-level search interest tracked survey-reported illness (r = 0.52)
  • Built an interactive US map, PCA and correlation views, and a plain-language assistant
Python SQL Pandas Chart.js
Tableau dashboard of mental-health risk indicators by countryMachine learning · Tableau

Mental Health Risk Indicators

Population-level analytics · machine learning · Tableau

Who seeks mental-health treatment among 1,258 surveyed tech workers, and how suicide rates in 185 countries changed from 2000 to 2021. The goal is evidence that can inform suicide prevention.

  • Combined the OSMI workplace survey with World Bank development indicators
  • Fit logistic regression, region-adjusted and within-country models, and cross-validated ML
  • Found family history is the strongest predictor of treatment (adjusted odds ratio 4.5)
  • Showed suicide rates fell 27% across 185 countries from 2000 to 2021 (World Bank / WHO data)
  • Published an interactive world map, Tableau dashboards and a plain-language assistant
Python scikit-learn Tableau Chart.js
Scatter plots of weather variables against latitudeAPI · visualization

Global Weather Analysis

Data visualization project

How does weather change closer to the equator? Live weather for 547 cities worldwide, collected and analyzed end to end.

  • Collected live weather data from the OpenWeatherMap API
  • Fit per-hemisphere regressions of four weather variables against latitude
  • Showed latitude explains most of the variation in temperature (r = −0.73 north, +0.83 south) but under 2% for humidity, cloud cover or wind
  • Charted the results with Matplotlib and Seaborn
Python API Matplotlib Seaborn
Charts comparing drug regimens and tumor volumeStatistics

Pharmaceutical Data Analysis

Statistical analysis project

A preclinical study of 249 mice with squamous cell carcinoma, comparing ten drug regimens with summary statistics, outlier tests and regression.

  • Cleaned the study data and removed a mouse with duplicate timepoints
  • Showed Capomulin and Ramicane left the smallest tumors (median ≈37 mm³ vs. ≈60 mm³ for Infubinol and Ceftamin)
  • Linked tumor volume to body weight under Capomulin (r = 0.84)
  • Flagged outliers with the IQR rule and charted results with Pandas and Matplotlib
Python Pandas Matplotlib Statistics
Sunken Garden at The Butchart Gardens, from the photography portfolioWeb app · personal

ARC-NATURE Photography

Photo platform · front-end engineering

My photography portfolio as a fast, installable web app: 1,000+ photos with search and categories, a travel map of every location, and a photography assistant.

  • Served responsive, lazy-loaded thumbnails for fast mobile browsing
  • Built an interactive Leaflet travel map from photo metadata
  • Added a retrieval-based assistant with calculators and gallery search
  • Enabled offline browsing with a service worker
JavaScript Leaflet PWA Accessibility

Professional journey

Work experience

  1. 2021 – Present

    Senior Analyst

    Translate complex collective-bargaining data into actionable insights through advanced analytics, research, AI-driven automation, and strategic reporting to support evidence-based decisions.

    • Research and analyze collective-bargaining trends to inform contributing to strategic decision-making
    • Implemented AI tools that extract and process data from bargaining documents, improving accuracy and efficiency
    • Design interactive dashboards and reports for internal and external audiences
    • Developed standardized methodologies for reporting key bargaining metrics
  2. 2018 – 2021

    Research Associate

    Conducted research and extracted business data to acquire new customers across a range of industries using internal tools and external data sources.

    • Built and enriched prospect lists in the company database
    • Cleaned, transformed and deduplicated records to resolve conflicts
    • Produced statistical reports and KPI dashboards
    • Documented acquisition strategies and identified process improvements
  3. 2011 – 2015

    Cyber Media Analyst

    Analyzed public complaints and social-media sentiment and advised government agencies on how to respond.

    • Monitored public inquiries across news outlets, government reports and social networks
    • Processed daily data from online forums, blogs and news sites
    • Wrote weekly reports on digital sentiment and public-opinion trends
    • Recommended responses to negative reviews and complaints

Academic career

Education

  1. 2024 – 2025

    Master of Science in Data Science

    Eastern University · St. Davids, Pennsylvania, USA

    Machine learning, advanced analytics (Python, R, SQL), database management, data visualization and statistical modeling, applying data science to business problems and building ethical, real-world AI solutions.

  2. 2019 – 2020

    Certificate in Data Analytics and Visualization

    University of Toronto School of Continuing Studies · Toronto, Canada

    A 24-week intensive boot camp in analyzing large datasets and communicating data-driven insights.

  3. 2015 – 2017

    Master of Science in Cognitive Science (Cognitive Neuroscience stream)

    CIMeC, University of Trento · Rovereto, Italy

    Research-focused master's in cognitive neuroscience.

  4. 2008 – 2011

    Bachelor of Arts in Psychology

    Hawassa University · Hawassa, Ethiopia

    Psychological sciences and techniques for analyzing and evaluating behavioral processes.

Let's work together

Get in touch

Open to data science roles, research collaborations and project inquiries.

Your details are used only to reply to your message and are never shared. Messages are delivered by Formspree.

Email · replies within 24 hoursermiasgelaye@gmail.com
LocationToronto, ON, Canada