Software, Data & Mathematics - Toronto, Canada
I'm a Computer Science graduate who enjoys building useful applications, finding patterns in data, and turning mathematical ideas into working code. My projects explore financial analysis and everyday workflows that can be improved through automation.
Current focus
In active development · updated September 2026
Machine-learning equity research · FastAPI
A research workstation that ranks 98 US large caps by expected sector-relative return, compares five models and an ensemble on embargoed walk-forward folds, and turns the signal into a cost-aware portfolio it keeps monitoring.
Twelve projects. Each one opens its own page with the full write-up.
Quantitative finance · Featured
A reproducible research system that turns ETF market data into quarterly portfolio weights, then tests those weights on five years of returns it never saw. SQL to pandas to a constrained optimizer, cross-checked in MATLAB and covered by tests.
Sharpe 0.79 vs 0.69 SPYWorst drawdown −18.4% vs −24.5%Beat all four naive allocation rules
Investment analytics · Data engineering
A Python and SQL pipeline that validates 19,575 historical price records, compares four Canadian equity portfolios, and extends completed backtests into five-year scenario ranges with explicit timing controls.
ML equity research · FastAPI
A machine-learning research workstation that ranks 98 US large caps by expected 20-session sector-relative return. Five models and an ensemble are tested on embargoed walk-forward folds against a momentum baseline, then turned into a cost-aware portfolio with live signals, per-stock attribution, and model monitoring.
Rank IC 0.024 t = 2.4, out of sampleNet Sharpe 1.24 vs 1.09 universe98 stocks · 6 models · live signal
IT service operations · SLA analytics
An analytics platform that monitors ticket volume, backlog, resolution time and SLA compliance across 50,000 simulated service desk tickets. Automated data-quality checks repair or quarantine duplicates, missing fields, inconsistent classifications and invalid timelines before reporting, and the analysis traces 61% of SLA breaches to two categories.
Breaches from 2 categories 61% on 34% of volumeSLA compliance 86.9% vs 90% goal15 quality checks · MySQL views · live dashboard
Real-estate portfolio risk
A property risk system that turns multifamily financials, CMHC-aware debt, valuations, and capital projects into stress tests, risk scores, concentration analysis, and budget-aware capital recommendations.
Cloud FinOps · AWS operations
An end-to-end, read-only AWS cost and resource monitoring platform that tracks spend, inventories infrastructure, detects idle and oversized resources, flags unusual cost spikes, estimates savings, and produces a prioritized action queue.
Independent research · AI economics
A 22-page scenario-based study comparing the operating cost of agentic AI with human labour across routine, tool-using, complex, and high-risk workflows. Includes break-even analysis and 11 evidence tables.
AI Connector
A private MCP connector that gives Claude verified, structured access to a property portfolio with eight read-only tools, deterministic business logic, and a full audit log. Deployed to AWS Lambda and passing all 32 evaluation checks.
Real-estate expense intelligence
A full-stack review workspace that turns property invoices into validated records, portfolio trends, explainable anomalies, and assigned follow-up actions. Live on Render with 565 demo invoices.
Mobile app
A mobile-first Python course built around short lessons, streaks, and instant feedback - programming practice shaped like a daily habit rather than a textbook.
Professional services · Applied AI
A partner prospecting and qualification workspace for professional-services teams. It ranks which reseller to call from renewal, end-of-support and whitespace signals, turns call notes into a BANT-qualified opportunity with the Claude API, routes work to specialists with a handoff brief, and drafts outreach that a person reviews before sending.
Site reliability · MLOps
An incident intelligence platform on simulated telemetry: anomaly detection with a measured precision vs. time-to-detect trade-off, an incident classifier that beats a runbook rules baseline and routes an unseen failure type to "unknown", SLO burn-rate alerts, a cited RAG assistant that refuses when evidence is missing, and drift-gated retraining with a deployment gate.
Technical delivery, data investigation, and customer-facing work in environments where accuracy and communication matter.
Volunteer Project Developer
Contribute to an MIT-affiliated agentic AI research project, applying analytical thinking and clear communication within a distributed technical team.
Asset Management & Risk Analyst Intern
Analyzed multifamily portfolio performance and financial risk, translating property-level operating data into practical decision support for asset management.
Data Analytics & Cloud Apprentice
Worked within the AbbVie client account, combining data analysis, automation, cloud training, and clear technical documentation in a security-conscious enterprise environment.
Customer Support Representative
Supported customers and high-volume store operations in a regulated environment where transaction accuracy, discretion, and calm problem resolution were essential.
Store Associate
Worked across a busy sales floor with frequently changing inventory, balancing customer support, product knowledge, and shifting operational priorities.
A coursework notebook exploring data cleaning, feature engineering, linear and regularized regression, and the bias-variance tradeoff using NYC Airbnb listings. Includes the original code, saved results, and visualizations.
View coursework notebookContinued learning in data, mathematics, AI, and cloud technologies, with viewable completion credentials.
01
MathWorks Training Services
Learning focus
Completed self-paced MATLAB training covering core syntax, arrays, plotting, scripts, functions, and practical numerical workflows used in technical computing.
02
IBM via Coursera
Learning focus
Completed IBM-authorized coursework introducing distributed data systems and the role of Spark and Hadoop in large-scale data processing.
03
AWS Training & Certification
Learning focus
Completed AWS training focused on the concepts and practical foundations behind agentic AI systems in the AWS ecosystem.
04
Amazon Web Services via Coursera
Learning focus
Built foundational understanding of AWS services, cloud architecture, security, pricing, and the shared responsibility model.
05
Microsoft via Coursera
Learning focus
Developed a foundation in Azure cloud concepts, core services, management tools, security, governance, and cloud-based solution planning.
06
Microsoft via Coursera
Learning focus
Strengthened cloud fundamentals across identity, access control, virtual networking, and secure Azure resource connectivity.
07
HCLTech
Learning focus
Completed an applied learning journey covering AWS delivery practices and the automation mindset used in modern DevOps teams.
The tools I use to build software, analyze data, implement mathematical models, and automate repeatable work.
Full-stack applications, APIs, desktop utilities, and practical automation tools.
Cleaning and validating data, building research pipelines, and communicating findings through clear visualizations.
Applying statistics, constrained optimization, and financial calculations in portfolio research, with explicit assumptions and reproducible tests.
Scripts and controlled integrations that reduce manual work and connect reliable data to useful interfaces.
Supporting applications with hosted services, repeatable deployments, and automated releases.
Investigation, monitoring, documentation, and communication that keep technical work understandable and supportable.
I'm drawn to the connection between software and mathematics: taking an idea, expressing it clearly, and building something I can test. I enjoy investigating data, understanding why a result changes, and making the reasoning easy for someone else to follow.
My portfolio research projects bring those interests together through financial calculations, optimization, and scenario analysis. My application and automation projects explore another side of the same curiosity: how to make repetitive work simpler and useful information easier to access.
I'm early in my career and eager to keep learning from experienced developers and analysts. I want to contribute careful code, thoughtful questions, and practical tools that help a team get things done.