Uncategorised

Details

Written by: Christopher Cooke

Category: Uncategorised

Published: 05 February 2026

Hits: 302

I'm a data scientist, astronomer, and chronic overthinker who's spent the last decade trying to figure out why people do what they do—not just what they do.

Most of my professional work has been in games and educational technology, building experiments, measurement systems, and models that help teams understand their users and make better decisions. But before I got into industry, I spent years studying stars, teaching physics, running a public observatory, and applying machine learning to clinical neuroimaging data. All of which is to say: I've been looking at complex systems and trying to make sense of noisy data for a very long time.

What I'm Actually Trying to Do

Here's the thing: I think my job is to help people understand things. Not to build models that make decisions for them. Not to automate away judgment. Not to produce dashboards that tell executives what they want to hear while obscuring what's actually happening.

I want to know why one feature gets ignored while another sees constant use. Why some users convert and others don't. What's broken, why it's broken, and whether we can actually fix it—or if we're just going to pretend the metric we're optimizing for is the thing that matters.

My role is to create knowledge that makes other people's work clearer and their decisions better informed. To translate what users are doing into insights about what they're trying to accomplish and where our products are failing them. To help designers, engineers, and product teams understand the gap between their intentions and what's actually happening in the wild.

But I'm increasingly convinced this puts me at odds with a lot of what "data science" has become. There's a whole contingent of the field that's really just building increasingly sophisticated ways to avoid thinking—to hand off responsibility to algorithms while maintaining plausible deniability. To produce analysis that looks rigorous but is actually just pandering bullshit designed to flatter management and confirm decisions they've already made.

I'm not interested in that. At all.

How I Think About Analysis

I approach data analysis as a process of knowledge creation, not pattern extraction—and I think the distinction matters more now than ever. We live in a world drowning in data but starving for understanding, where algorithms claim to "let the data speak for itself" while actually encoding the assumptions and incentives of whoever built them. Data doesn't explain behavior on its own. It reflects the interaction between people, systems, and the ideas or incentives that shape their choices. If you don't understand the system, you're just finding patterns in noise.

This means meaningful analysis starts with theory—with domain knowledge, an understanding of human motivation, and a clear mental model of what's actually happening. Before I ask "what are users doing?" I ask "who are these people, and what constraints are they operating under?" Segmentation isn't just a clustering exercise; it's a hypothesis about how the world works. Groups aren't accidents in the data—they're populations bound by shared goals, constraints, or contexts. Whether I'm analyzing players, learners, research subjects, or customers, I treat defining these groups as an act of scientific inquiry, not statistical convenience.

From there, I focus on relationships: between who people are, what they do, and what happens as a result. Funnels, retention curves, and predictive models are tools for surfacing these relationships, but they're worthless without context. I want to know what motivates people, what trade-offs they face, and how the structure of the product or environment shapes their decisions. This perspective draws as much from psychology and domain expertise as from statistics—because understanding behavior requires understanding people, not just their clicks.

I'm particularly skeptical of claims about causality that aren't grounded in experimentation or counterfactual reasoning. Observation alone is insufficient. You need experiments, careful metric design, and an honest accounting of what you can and can't conclude from your data. Even well-designed experiments only tell you part of the story—causality emerges most clearly when you reconcile quantitative results with qualitative knowledge of the system, its mechanics, incentives, and failure modes.

This philosophy has been consistent across everything I've done: astrophysical simulation, medical machine learning, live service games, educational products. The contexts change, but the goal stays the same: build models and analyses that are not only statistically sound, but conceptually honest. Grounded in how real people engage with real systems. Capable of supporting decisions that actually matter.

I care about this because data science done poorly is just sophisticated bullshit. And too often, it's pandering bullshit—designed to flatter management, feed egos, and create the illusion of rigor while avoiding any actual insight that might be uncomfortable or inconvenient. Done well, though, it's a tool for building better understanding and making better decisions. That's what I'm trying to do.

What I Care About

What drives me is the same thing that got me into science in the first place: I want to understand how things work, and I want to help other people understand them too.

I'm drawn to problems at the intersection of human behavior, system design, and measurement—places where understanding the structure of the system is as important as analyzing its data. I like work that requires both curiosity and rigor, where you need to know when to trust your intuition and when to demand proof, and where the goal is to enable better decisions rather than automate them away.

I also care deeply about making space for different ways of engaging with systems. Whether it's pushing back on the idea that there's only one "right" way to play a game, or questioning whether optimization is actually the point, or just refusing to accept that complex tools can only be used in rigid ways—I think a lot of received wisdom in technical communities is less about what's true and more about what's comfortable for the people saying it. And I'm not particularly interested in being comfortable if it means pretending everyone has to care about the same things I do.

As I look for my next role, I'm looking for teams that value understanding over automation. That want analysts who ask hard questions, not ones who produce pretty charts that confirm existing biases. That see human judgment as essential, not something to be optimized away.

Outside Work (Or: The Things That Actually Define Me)

In a lot of ways, I think the things I care about outside of work say more about who I am than anything in my professional life.

Astronomy has been part of me for as long as I can remember. I grew up watching the Moon from my bedroom window, seeing comets dominate the night sky, gazing at the winter Milky Way from my snowy backyard in rural Atlantic Canada. I studied it in university, did my master's degree in it, and spent years running one of Canada's largest planetariums—showing thousands of people the stars overhead, the planets, the galaxies, everything we can observe in the universe above. I got to do real scientific research, operate observatory telescopes, give public tours. My dream was always to be a professional astronomer. That didn't work out—I didn't get selected for a PhD at my school, and I chose to stay home and make a life here rather than chase that dream somewhere else. It's one of my deepest sadnesses, but also one of the things that made me who I am. I still stare at the sky whenever I get the chance. Once an astronomer, always an astronomer.

Baseball is, in a very real and visceral way, who I am. I've been playing in some capacity since I could walk—whether that's by myself in the backyard, in little league as a pitcher, in video games, or now in recreational softball 2-3 times a week every summer. I still remember clearly (and occasionally rewatch) the Blue Jays' back-to-back World Series victories in 1992 and 1993. I follow the Jays as much as I can. There's something about the rhythm of the game, the elegance of the mechanics, the way it unfolds that just speaks to me in ways nothing else quite does.

Star Trek -- specifically 90s Trek: The Next Generation, Deep Space Nine, and yes, even Voyager -- formed the very core of my personal politics. These shows, TNG especially, taught me that humanity deserves to be free, and that we can only be free if we're not strapped to the yoke of unchecked materialism and unjust, or non-consensual, power hierarchies. It's idealistic, sure, but it's the kind of idealism that asks: What if we actually believed people were worth investing in? What if we built systems that served human flourishing instead of extraction? Those questions shaped how I think about everything, including what kind of work I'm willing to do and what kind of world I want to help build.

Video games have been my constant companions since I was maybe four years old, playing the original Legend of Zelda on the NES. I've been a lifelong Zelda fan—A Link to the Past and Ocarina of Time are games I've replayed so many times I've lost count. The original Super Mario Bros., Mario 2, Mario 3, Super Mario World—I've played them so much I probably dream about them. The Super Mario Bros. 2 overworld theme still gets stuck in my head 35 years later. And Mario Kart? Easily my favorite franchise of all time. I bought a Wii U expressly to play Mario Kart 8. I've probably logged more hours into Super Mario Kart than any other game ever. Beyond Nintendo, I have deep, formative memories of Final Fantasy VII (which I'm replaying for the first time in 25 years and loving just as much as I did in junior high), point-and-click adventure games (probably my most beloved genre conceptually, even if I haven't played that many), and the mid-90s FPS trinity of Doom, Doom II, and Duke Nukem 3D—games I used to play in the computer lab with friends, in dial-up deathmatches, and that I still return to 30 years later.

I also GM TTRPG sessions for my friends and family, Pathfinder 2e specifically. My home game is set in a homebrew world (Ardrigail) based on Hyrule from Ocarina of Time. I love the system, but I find myself sharply at odds with a lot of the online community that treats it as a combat engine first and a roleplaying game last. To me, it's a rich and reliable toolkit for collaborative storytelling that happens to have robust tactical rules when you need them, not the other way around. The prescriptivism and gatekeeping I see in a lot of TTRPG spaces frustrates me endlessly, which is probably why I keep writing about it.

All of this -- the Star Trek, the games, the baseball, the stars -- feeds into how I think about my work. I'm constantly asking: What are the systems at play here? What are people trying to accomplish? Where does intention meet reality, and what happens in that gap? Whether I'm analyzing user behavior in a game or thinking about why someone swung at a pitch outside the zone or pondering how light from a distant galaxy reached my eye, it's all the same curiosity. The same need to understand.

Details

Written by: Christopher Cooke

Category: Uncategorised

Published: 05 February 2026

Hits: 256

For over a decade now, I’ve had the pleasure – and the frustration – of being a data scientist in the video games industry. I’ve always loved problem-solving. I’ve always loved video games, computers, and digital products of all kinds (okay, most kinds). They entertained me, fascinated me, and gave me worlds to explore in ways that the people around me never really seemed to care about when I was growing up.

I’ve also always loved science – all science, though none more than astronomy. Getting to use the scientific method, my problem-solving instincts, my personal understanding of digital products, and my technical skills to help create better video games has been a huge privilege. One I never could have imagined when growing up in rural Atlantic Canada.

I’ve had the opportunity to work with some truly amazing people – including many I never quite managed to form meaningful personal connections with. If any of my former colleagues at Ubisoft or Prodigy are reading this, please know that you had a profound impact on me. I’m deeply grateful to have crossed paths with you, even if my quiet, awkward self never quite figured out how to fully engage. Working alongside artists, designers, engineers, producers, and monetization teams gave me years of chances to reflect on how my own profession can – and should – intersect with theirs. In many cases, perhaps too late, I learned how data could genuinely support their work and, I hope, make their jobs a little easier.

Along the way, though, I’ve also encountered some… uncomfortable ideas about what data science and analytics are supposed to be. And more often than not, those ideas came from within my own field. I’ve come to realize that my goals as a data scientist often diverge sharply from the goals of many others who share my title. That the “science” in “data science” is, for some, little more than a word of distinction, rather than a signal of methodology, ethos, or intent. I’ll admit: when that finally sank in, it was a pretty profound shock.

There are plenty of people who see data science as little more than the automation wing of software engineering – marching steadily toward a future where decision-making itself is automated away, responsibility is handed off to an inscrutable black box, and human judgment quietly exits the room.

Man, oh man, is that terrifying. And the thing is, it’s the natural endpoint of organizations that proudly embrace the phrase “data-driven decision making.”

After all, if the data is the one doing the driving, then you’re just a passenger.

I find that vision bleak, and now that I’m looking for my next role, it’s one I know I need to be careful not to stumble into.

See, I’ve always thought of my job as being that of a knowledge facilitator, not a process automator. As someone who works to understand what my colleagues are trying to accomplish, what problems they’re actually trying to solve, and then uses that context (along with my own experience) to understand how real people engage with our products. To try to understand what those users’ motivations and goals are, what the designer’s intentions are, and then to understand where real behaviour diverges from those, and why.

 

I want to understand everything, always. That includes why one part of an application is heavily used while another is ignored. Why one user makes a purchase and another walks away. What isn’t working, why it isn’t working, and what – if anything – we can do about it.

 

I desperately want to understand it all. Everything else follows from that.

And I want to bring that understanding to others. Not to replace their judgment, but to inform it. To create knowledge that makes people’s work a little easier, their decisions a little clearer, and – if we’re lucky – the products we build a little better for the people who use them.

I worry, though, that this isn’t where my field is headed. That too many businesses aren’t run by people who value understanding, but by people eager to offload responsibility – while clinging to the illusion of control.

More and more, I worry that we’re becoming comfortable being passengers in our own vehicles.

Details

Written by: Christopher Cooke

Category: Uncategorised

Published: 29 January 2026

Hits: 336

Table of Contents

• Professional Summary
• Analytical & Research Philosophy 
• Skills, Tools & Methods
• Professional Experience
• Education
• Publications

Professional Summary

I am a Senior Product Data Scientist and astronomer with over a decade of experience designing experiments, building measurement systems, and translating complex questions into rigorous analysis. My professional work spans live service games, educational technology, and applied machine learning research, but my intellectual foundation is in astronomy and physics—fields that taught me how to work with noisy data, validate models against reality, and communicate difficult concepts to diverse audiences.

I specialize in the full analytics lifecycle: designing telemetry and event instrumentation for complex systems, owning end-to-end A/B testing from metric definition through causal analysis, building predictive and behavioural models, and constructing analytics infrastructure that teams can trust. My approach emphasizes building durable foundations -- measurement frameworks, experiment platforms, and data pipelines -- that enable others to ask better questions and make informed decisions.

Across education technology and consumer products, I have led experimentation for product decisions delivering measurable growth in subscription conversion, user engagement, and revenue. I have designed analytics-ready datasets in Databricks and dbt, built self-service dashboards, and partnered directly with executive leadership during organizational transitions. I have also contributed to peer-reviewed research in applied machine learning for medical diagnostics and conducted original computational research in stellar atmospheres.

My background as an astronomer, educator, and published researcher informs how I approach data science: careful attention to assumptions and uncertainty, comfort working with incomplete or messy data, and a commitment to making complex findings accessible without sacrificing rigour. I spent five years operating a public planetarium and observatory, translating the physics of stars, galaxies, and planetary motion into narratives that audiences could understand and connect with—skills that transfer directly to translating user behaviour and experimental results into actionable insights for product teams.

I treat analytics as a process of knowledge creation, not pattern extraction. I believe the best analysis combines statistical rigour with domain expertise and an understanding of the systems and incentives that shape behaviour. I am drawn to problems at the intersection of human behaviour, system design, and measurement, where understanding the structure of the system is as important as analyzing its data.

Analytical & Research Philosophy

I approach data analysis as a process of knowledge creation, not pattern extraction. Data does not explain behaviour on its own; it reflects the interaction between people, systems, and the ideas or incentives that shape their choices. Meaningful analysis therefore begins with theory: domain knowledge, an understanding of human motivation, and a clear mental model of the system being observed.

Across product analytics, applied machine learning, and scientific research, my work centers on identifying who is being observed before asking what they are doing. This means defining and discovering group membership—sometimes explicit and self-selected, sometimes implicit and behavioural, and sometimes constructed through careful analysis. Whether these groups are users, learners, players, or research subjects, I treat segmentation as a hypothesis-driven act: groups are not merely clusters in data, but populations bound by shared constraints, goals, or contexts.

From there, I focus on the relationships between group membership, behaviour, and outcomes. Funnels, retention curves, and predictive models are tools for surfacing these relationships, but they are always interpreted within a broader scaffold of understanding: what motivates these groups, what trade-offs they face, and how the structure of the product or environment shapes their decisions. This perspective is informed as much by psychology and domain expertise as by statistics.

I view causal understanding as something that must be actively constructed. Observation alone is insufficient; experiments, counterfactual reasoning, and careful metric design are necessary to separate signal from artifact. Even then, causality emerges most clearly when quantitative results are reconciled with qualitative knowledge of the system—its mechanics, incentives, and failure modes.

This philosophy underpins my work from astrophysical simulation and medical machine learning to large-scale product experimentation. In every context, my goal is the same: to build models and analyses that are not only statistically sound, but conceptually honest—grounded in how real people engage with real systems, and capable of supporting decisions that matter.

Skills, Tools & Methods

Experimentation & Causal Inference A/B Testing & Multivariate Testing: Full lifecycle ownership from experiment scoping and metric definition through power analysis, monitoring, evaluation, and stakeholder communication Causal Reasoning: Hypothesis testing, pre/post analysis, causal interpretation with attention to confounding and selection bias, guardrail metrics, counterfactual thinking Measurement Design: Event instrumentation and telemetry architecture, KPI frameworks, success criteria definition, outcome measurement and forecasting Analytics Engineering & Infrastructure Data Modelling: Analytics-ready datasets in Databricks and dbt, ELT workflow design, reproducible analysis pipelines, automated reporting systems Data Quality & Reliability: Telemetry validation, lineage tracing, integrity checks, anomaly detection, production monitoring Platforms & Tools: Databricks, dbt, PySpark, Git, cloud analytics environments Leadership & Communication Team Leadership: Shared leadership, mentorship, priority coordination, task delegation, analytics enablement and training Stakeholder Partnership: Cross-functional collaboration with Product, Engineering, Data Science, Finance, and Executive teams Communication: Technical writing, executive presentations, data storytelling, scientific communication, public science education, teaching and lecturing

Statistical Modelling & Machine Learning Statistical Methods: Multivariate regression (linear and logistic), cluster analysis, curve fitting, time series and trend forecasting Applied Machine Learning: Support Vector Machines, Gaussian Process Classifiers, classification and predictive modelling in high-dimensional, noisy, or constrained data regimes Model Validation & Evaluation: Cross-validation strategies, significance testing, error analysis, interpretability assessment, robustness checks Behavioural Analysis & User Lifecycle Funnel & Journey Analysis: Onboarding and activation funnels, feature adoption tracking, user pathing, conversion optimization, friction and drop-off analysis Retention & Engagement: Churn prediction models, engagement scoring, cohort analysis, lifecycle segmentation, long-term trend forecasting Segmentation & Pattern Detection: Behavioural clustering, market basket and association analysis, ratio-based feature engineering Technical Foundations Python: Pandas, NumPy, SciPy, Statsmodels, Scikit-learn, PySpark, Matplotlib SQL: Advanced querying, window functions, CTEs, analytics engineering, performance optimization Julia: DataFrames, GLM, Gadfly/Plots Visualization & Analytics Enablement Dashboards & Reporting: Stakeholder-focused visualization using Tableau, Databricks Dashboards, Periscope Data, Superset, Matplotlib, Excel Self-Service Analytics: Parameterized dashboards and tools enabling independent exploration and data-driven decision-making by non-technical stakeholders

Professional Experience

Prodigy Education

Toronto, ON (Remote)

Data Scientist II | Jan 2024 – Jan 2026

Joined as a senior-level analytics hire with 5 years of game analytics experience, immediately taking ownership of experimentation, telemetry architecture, and product decision support for an educational game serving millions of monthly users.

• Experimentation & Product Decisions: Designed and analyzed A/B tests for 10+ major product releases, providing experimental evidence and launch recommendations for features that delivered +15% subscription conversion, +28% growth in high-value players, and +22% revenue per install
• Economy & Engagement Analysis: Quantified behavioural impact of game economy changes, informing system designs that increased premium currency participation by 14%, play intensity by 11%, and early-stage progression by 6%
• Leadership & Team Management: Co-led 5-person analytics team (2 senior data scientists, 3 analysts) following Director departure; owned onboarding, mentorship, project reviews, and task prioritization for 3 new hires while advocating for their autonomy with cross-functional stakeholders
• Executive Partnership: Maintained direct partnership with CFO on team priorities, project roadmaps, and organizational dynamics during 6-month leadership transition
• Rapid Response Analytics: Led forensic analysis during unexpected revenue decline, isolated faulty feature interaction within 48 hours, and enabled corrective release that stabilized sales
• Telemetry Architecture & Measurement Systems: Owned redesign of product telemetry for interconnected game systems; established metrics definitions and experiment frameworks during migration from Optimizely to GrowthBook; trained product managers on telemetry best practices
• Data Engineering & Reliability: Developed PySpark lineage tooling that enabled detection of invalid accounts, restoring metric integrity; designed dbt data models consolidating disparate sources into analytics-ready datasets
• Predictive Modelling & Behavioural Insight: Built churn prediction models, analyzed onboarding and progression funnels to identify friction points, and designed self-service A/B test dashboards enabling teams to monitor experiments independently
• User Journey Analysis: Mapped complex user pathing through interconnected game environments, identifying navigation choke points and layout issues associated with session termination and player drop-off

Data Scientist I | Dec 2022 – Dec 2023

• Behavioural Modelling: Built multivariate linear and logistic regression models in Databricks to quantify trends in user behaviour, learning outcomes, and feature engagement
• Experimentation Support: Partnered with product stakeholders to design A/B tests, define success metrics, and analyze experimental results, supporting data-driven feature and experience decisions across the product organization
• Analytics Engineering: Leveraged dbt to create integrated, analytics-ready database views, consolidating disparate data sources and enabling consistent metric definitions across the analytics organization
• Self-Service Analytics: Developed and optimized generalized A/B testing dashboards, enabling stakeholders to monitor experiment performance independently and reducing ad-hoc reporting requests
• Platform Evaluation: Evaluated and piloted business intelligence and data visualization tools to inform analytics platform strategy and tooling investment decisions
• Analytics Enablement: Trained product managers on designing robust telemetry events and instrumentation strategies to ensure high-quality data capture for new feature launches

Data Analyst, Game | Mar 2021 – Nov 2022

• Exploratory Analysis: Led exploratory analyses of the onboarding funnel, feature engagement patterns, and correlates of retention and conversion, translating early performance signals into actionable product insights
• Telemetry Strategy & Redesign: Served as subject-matter expert for video game and freemium application data collection; redesigned product telemetry to capture dependencies and interactions between interconnected game systems, enabling richer behavioural analysis and more effective experimentation
• Stakeholder Partnership: Partnered with product owners and cross-functional teams to ensure accessibility, clarity, and usability of dashboards, reports, and experiment results; translated analytical findings into non-technical recommendations
• Analytics Training: Increased product owner self-sufficiency in data retrieval and interpretation by delivering hands-on SQL training sessions and creating concise, reusable reference materials
• Cross-Functional Advisory: Advised analytics groups and cross-functional teams on the use, interpretation, and limitations of game-specific user data in decision-making contexts

Ubisoft Halifax

Halifax, NS

Analyst, Consumer & Market Intelligence | Apr 2019 – Feb 2021

Led monetization analysis, player segmentation, and analytics infrastructure for live-service mobile titles.

• Stakeholder Communication: Communicated analytical findings through written reports and presentations to designers, producers, and cross-functional stakeholders, translating complex statistical results into actionable product recommendations
• Technical Mentorship: Acted as senior analytical resource, mentoring and training junior analysts and interns on Python, SQL, and analytical best practices

Assassin's Creed: Rebellion (Platformer RPG & Base Builder)

• Telemetry Redesign: Completely redesigned event instrumentation and tracking architecture for live product; built comprehensive dashboard suite that remained in production use through studio closure (5+ years after departure)
• Live Operations Analysis: Used trend analysis and pre/post statistical methods to evaluate performance of live events, content updates, and seasonal campaigns; informed live ops strategy and resource allocation decisions
• Monetization & Funnel Analysis: Analyzed in-game store sales performance and pricing strategies; evaluated user progression funnels to identify monetization opportunities and friction points

Tom Clancy's Rainbow Six Mobile (Pre-production; 5 v 5 asymmetrical team FPS)

• Telemetry Design: Designed initial event instrumentation and tracking strategy for pre-production build 

Data Analyst | Nov 2015 – Mar 2019

Led game and player behaviour analytics, developing reporting strategies and analytics infrastructure across multiple titles at various stages of the product lifecycle.

• KPI Infrastructure & Dashboards: Designed and automated daily KPI reporting and alerting pipelines in Excel for senior leadership, providing early visibility into performance regressions, growth trends, and critical business metrics across all supported titles

Tom Clancy's ShadowBreak (Soft launch - PvP sniper shooter)

• Analytics Architecture: Designed comprehensive telemetry strategy and dashboard infrastructure from pre-launch through soft launch operations
• Forecasting & Capacity Planning: Applied ARIMA models and Singular Spectrum Analysis to predict future daily active users and sales trends; used survival analysis to estimate concurrent user loads, informing server infrastructure and capacity planning decisions
• Player Segmentation & Outlier Detection: Conducted cluster analysis to identify engagement and spending profiles; implemented outlier detection to identify potential cheating or exploitation patterns
• Metrics Definition & Partnership: Partnered with product designers, economy and monetization specialists,and marketing stakeholders to define success metrics, evaluate feature engagement, and assess player performance across releases 

Rock Gods (Soft launch - ad-monetized idle clicker)

• Early-Stage Analytics: Designed telemetry and built dashboards for ad-monetized idle game during soft launch evaluation period
• Dashboard Engineering: Built and maintained comprehensive Tableau dashboards tracking feature adoption, engagement, retention, and long-term game health for senior leadership and product teams

Independent Contractor

Halifax, NS

Data Analyst | Jan 2014 – Nov 2015

• Performed statistical analysis and data modelling across diverse applied projects, including spam detection algorithms and sports analytics pipelines
• Developed custom algorithms to map GPS trajectory data onto digital image maps, supporting spatial analysis and data visualization use cases
• Produced research and analysis reports for non-technical audiences, including step-by-step explanations and worked examples to support adoption of developed analytical methods

Social Navigator, Inc.

Halifax, NS

Research Consultant | Sep 2014 – Jan 2015

• Designed and executed experiments to evaluate the performance, efficiency, and suitability of third-party computing services and applications for integration into new and existing products
• Built MVP for a facial recognition and landmark-based geolocation feature using OpenCV in Python, supporting early-stage feature exploration and technical feasibility assessment
• Analyzed experimental results using statistical methods across R, MATLAB, Python, and Excel to compare trade-offs, performance characteristics, and fit-for-purpose criteria
• Authored detailed research reports documenting methodology, findings, and recommendations, supporting product strategy and technical integration decisions for executive stakeholders

Dalhousie University, Department of Psychiatry

Halifax, NS

Research Assistant | Nov 2013 – Jun 2014

• Evaluated the viability of machine learning methods, including Support Vector Machines and Gaussian Process Classifiers, for diagnostic applications in mental health research using structural MRI data
• Analyzed experimental data and performed hypothesis testing, cross-validation, and statistical significance analysis using MATLAB and specialized neuroimaging toolsets
• Implemented and automated third-party analysis pipelines, improving research efficiency, reproducibility, and experimental throughput
• Developed and optimized end-to-end project workflows to support iterative model development, validation, and performance evaluation
• Contributed to research accepted for publication in the Journal of Psychiatry & Neuroscience (2015)

Saint Mary's University, Department of Astronomy & Physics

Halifax, NS

Observatory Operator | Aug 2015 – Mar 2020

• Operated the Burke-Gaffney Observatory telescope and associated computing systems during weekly public observing sessions, serving diverse community audiences
• Selected and presented astronomical objects of interest based on seasonal visibility, current events, and audience composition; explained observational features and underlying physical principles in accessible language
• Guided guests in locating observed objects in the night sky, connecting telescope views to broader celestial context and naked-eye navigation techniques
• Communicated concepts in astronomy, astrophysics, and observatory technology in an engaging and accessible manner, ensuring positive and educational visitor experiences

Lecturer | Jan 2017 – Apr 2017

• Taught undergraduate physics courses in electromagnetism, delivering lectures aligned with departmental curriculum and learning objectives
• Developed instructional materials, problem sets, and assessments to support student understanding of core concepts in classical electromagnetism
• Coordinated with teaching assistants to ensure consistency and alignment between lecture content and laboratory coursework

Laboratory Instructor | Sep 2014 – Apr 2015

• Led weekly undergraduate physics laboratory sections of 15–30 students, ensuring clarity of experimental objectives, procedures, and safety protocols across diverse student backgrounds and experience levels
• Evaluated student lab reports and practical work, providing structured feedback and assigning grades in alignment with course standards and learning outcomes
• Coordinated and supervised teaching assistants to ensure consistent instruction, assessment quality, and student support across multiple lab sections

Research Assistant | Sep 2011 – Dec 2013

• Conducted original computational astrophysics research on the use of detailed stellar atmosphere models as upper boundary conditions for stellar interior and asteroseismology calculations
• Developed custom analysis and simulation tools in Python and Fortran to clean, reduce, and analyze 140+ GB of stellar atmosphere data, supporting model validation and parameter space exploration
• Modeled and interpreted simulation results, presenting findings through clear visualizations, tables, and written summaries for both technical and non-technical audiences
• Maintained detailed project documentation, including experimental goals, methodologies, intermediate results, and validation procedures
• Reviewed and synthesized technical literature across stellar physics, radiative transfer, and numerical methods; prepared public-facing presentations on research findings for departmental colloquia and public outreach events

Teaching Assistant | Sep 2011 – Apr 2013

• Led weekly tutorial and seminar sections supporting student skill development and conceptual understanding in undergraduate physics and astronomy courses
• Demonstrated laboratory techniques and assisted instructors during lectures and practical sessions, providing real-time support for student learning
• Evaluated student assignments, lab reports, and examinations; provided constructive feedback to support academic development
• Identified strengths and gaps in student support programs and communicated findings to course instructors and departmental administrators to inform curriculum improvement
• Participated in regular discussions with faculty and the Assistant Dean of Science to improve instructional effectiveness and student outcomes

TELUS World of Science

Edmonton, AB

Planetarium & Science Presenter | Apr 2007 – Aug 2011

• Wrote, scripted, and delivered live planetarium and stage shows for audiences of 300+, translating complex concepts in astronomy, physics, and Earth science into engaging, age-appropriate educational narratives for diverse audiences
• Performed interactive science demonstrations illustrating fundamental principles in physics, optics, motion, and energy; taught specialized classes on observational astronomy and the use of planetarium software
• Interpreted and presented permanent and traveling science exhibits; served as guide for public observatory tours and special programming events
• Trained colleagues in the operation of specialized planetarium projection hardware, presentation software, and digital control systems
• Installed, maintained, and performed minor repairs on presentation computers and audiovisual equipment to ensure reliable delivery of live public programming with zero downtime
• Researched current scientific literature and interviewed subject-matter experts across astronomy, physics, climatology, and planetary science to ensure accuracy, currency, and scientific integrity of all public-facing content

Education

Master of Science in Computational Astrophysics | 2014
Saint Mary's University, Halifax, NS

Bachelor of Science (Honours) in Physics | 2006
Queen's University, Kingston, ON

Publications & Research

Research Approach

My research work—spanning computational astrophysics, applied machine learning, and psychiatric neuroimaging—has always focused on building and validating models in settings where data are noisy, high-dimensional, or structurally constrained, and where getting the assumptions wrong has real consequences.

In astrophysics, this meant building physically grounded simulation pipelines and understanding how numerical approximations propagate through predictions. In clinical neuroimaging, it meant applying machine learning with rigorous validation and interpretability requirements in a high-stakes diagnostic context. Both required the same core skills: careful experimental design, honest error analysis, attention to what you can and can't conclude from your data, and translating complex technical work into results that others can trust and act on.

These habits carry directly into my industry work: designing experiments that actually answer the question, understanding where models fail, quantifying uncertainty, and communicating limitations as clearly as findings.

Upper Boundary Conditions for Asteroseismological Modelling of Solar-type Stars

Christopher Cooke, MSc Thesis, Saint Mary's University, 2013

Built and validated a computational grid of high-resolution stellar atmosphere models to improve outer boundary conditions in asteroseismology calculations. Using the PHOENIX radiative transfer code, I systematically evaluated interpolation strategies across temperature and surface gravity parameter space, quantifying how numerical choices (grid resolution, LTE vs. NLTE treatment) propagate into predicted stellar oscillation frequencies.

The work required designing experiments over high-dimensional parameter spaces, validating numerical stability, and translating astrophysical theory into scalable computation—skills that map directly to surrogate modelling, simulation-informed machine learning, and reliability analysis in production systems.

Methods: Radiative transfer modelling, high-dimensional interpolation, numerical error analysis, validation against observational constraints

---

Using Structural MRI to Identify Individuals at Genetic Risk for Bipolar Disorders: A Two-Cohort Machine Learning Study

T. Hajek, C. Cooke, M. Kopecek, T. Novak, C. Höschl, M. Alda
Journal of Psychiatry & Neuroscience, 2015

Applied supervised machine learning (Support Vector Machines, Gaussian Process Classifiers) to structural MRI data from two independent international cohorts to identify individuals at genetic risk for bipolar disorder before clinical onset. Achieved above-chance classification accuracy while maintaining interpretability through discriminative weight maps identifying distributed neuroanatomical patterns.

The study required working under tight constraints: small sample sizes, high dimensionality, site heterogeneity, and the need for both statistical rigor and clinical interpretability. Success depended on rigorous cross-validation, permutation testing, and close collaboration across disciplines—exactly the combination of technical depth and stakeholder partnership required for high-stakes applied data science.

Methods: Support Vector Machines, Gaussian Process Classifiers, cross-validation, permutation testing, neuroimaging analysis