Owen Dinsmore — Data Science & ML
Owen
Dinsmore
Building machine-learning systems for risk and markets. Boston University '27. Two summers in credit analytics at RenaissanceRe.
Get in touch LinkedIn ↗︎ GitHub ↗︎
BORN & RAISED — BERMUDA 32.3°N✳HIGHSCHOOL — TORONTO 43.7°N✳UNIVERSITY — BOSTON 42.4°N✳WORKING — NEW YORK 40.7°N✳AMERICAN · BERMUDIAN (PRC) · CANADIAN✳
BORN & RAISED — BERMUDA 32.3°N✳HIGHSCHOOL — TORONTO 43.7°N✳UNIVERSITY — BOSTON 42.4°N✳WORKING — NEW YORK 40.7°N✳AMERICAN · BERMUDIAN (PRC) · CANADIAN✳
NOW
B.S. Data Science at Boston University, May 2027. Dean's List. Investment Club quant research, men's rugby.
LATELY
Three summers in credit analytics at RenaissanceRe — fell in love with global credit risk trend analysis, AWS data pipelines, and reporting automation.
FOCUS
Machine learning, risk modeling, alternative data. Python, Rust, SQL — and building models from scratch to understand them.
01

Experience

2025 – present
Founder
StutterStep, Inc.
Founded and building StutterStep.
2025 – present
Lead Maintainer — Riskplot & ml-xray
Open source, PyPI
Lead maintainer of two published Python packages: Riskplot (risk visualization) and ml-xray (model inspection).
2024 – 2026
Credit Analytics Intern
RenaissanceRe
Backfilled AWS with credit data via API, analyzed global credit-rating trends, and returned to expand risk coverage — visualization tooling, automated report generation, SQL and Tableau.
2023 – 2025
Quant Research, Senior Developer
BU Investment Club
Reproduced a Fama-French 3-factor model predicting S&P 500 returns and led the portfolio-optimization group to redistribute club portfolio risk.
2021 – 2023
Researcher & Ambassador
MIT / Stanford Inspirit AI
Researched NLP for predicting market-sector trends from Twitter sentiment, advised by MIT and Stanford graduate students; invited back as ambassador.
2019 – 2023
Founder & President
Appleby College VEX Robotics
Founded and funded the school's competitive robotics program; took the team from non-competitive to #2 high school in Ontario across 11 competitions.
02

Selected projects

PACKAGE WEBSITE APP STORE LIVE
Perspective Data
2026
DATA · PLATFORM
Data platform project — description coming.
CLICK FOR MORE ↗︎
Prediction Market Stat Arb
2026
PYTHON · STAT ARB
Statistical arbitrage across prediction markets — pricing discrepancies between correlated contracts.
CLICK FOR MORE ↗︎
Bermuda Weather
2026
DATA · FORECASTING
Weather data and forecasting for Bermuda.
CLICK FOR MORE ↗︎
TerraGeo
2026
GEOSPATIAL
Geospatial tooling — description coming.
CLICK FOR MORE ↗︎
statsplayground
2026
STATISTICS · INTERACTIVE
Interactive playground for exploring statistical concepts.
CLICK FOR MORE ↗︎
ML-Xray Package
2026
PYTHON · ML · PYPI
Python package for inspecting models across pre-training and post-training — surfacing internals to diagnose behavior at each stage.
CLICK FOR MORE ↗︎
Stutter Detection
2025
PYTHON · ML · AUDIO
Machine-learning model for detecting stutter events in speech.
CLICK FOR MORE ↗︎
Multi-Factor Regime Detection
2025
PYTHON · HMM · BACKTESTING
Trading system that detects bullish, bearish, and neutral market regimes with Hidden Markov Models and switches strategies per regime.
CLICK FOR MORE ↗︎
Loan Default Prediction in Rust
2025
RUST · NDARRAY · ML
Logistic regression built from scratch with only ndarray, predicting HMEQ mortgage defaults at 70–80% accuracy.
CLICK FOR MORE ↗︎
RiskPlot Package
2025
PYTHON · MATPLOTLIB · PYPI
Python package extending matplotlib with publication-ready ridge plots for ordered qualitative distributions, published to PyPI.
CLICK FOR MORE ↗︎
Virtual Server Infrastructure
2025
LINUX · POSTGRES · REDIS
Deployed a stable remote Linux server for compute and storage with PostgreSQL, Redis caching and queues, and RESTful APIs.
CLICK FOR MORE ↗︎
Industry-to-Vec
2024
PYTHON · NLP · GLOVE
Vectorized NAICS industry codes with GloVe embeddings to compare similar industries, with utilities for embedding generation.
CLICK FOR MORE ↗︎
Twitter NLP for Trading
2023
PYTHON · NLP · SENTIMENT
NLP model reading Twitter sentiment to predict market-sector trends — research advised by MIT and Stanford graduate students.
CLICK FOR MORE ↗︎
RL Bitcoin Trading Bot
2023
PYTHON · RL · CRYPTO
Reinforcement-learning agent trained to trade Bitcoin, learning entry and exit policy from market state.
CLICK FOR MORE ↗︎
QUANT MODELS
Contrarian Reversal
Mean-reversion signal on short-horizon overreaction
CAPM & FF3 Optimization
Factor-model portfolio optimization
Senator Trading
Signal from congressional trading disclosures
News Sentiment
NLP sentiment signal from market news flow
Markov Country Prediction
Macro-driven Markov model allocating into country markets by region, macro effect, and country
+
More coming soon
03

Toolkit

LANGUAGES
Python Rust SQL Java C++ C#
DATA & ML
Machine Learning Pandas Scikit-Learn PowerBI Excel
PYPI
NumPy Pandas Scikit-Learn PyTorch statsmodels schwab-py duckdb
INFRA
AWS Linux PostgreSQL Redis
OFF DUTY
Rugby · Scuba diving · Spearfishing · Strength training · Travel · Investing · Computer hardware · Remote camping photography
04

Games, trained with RL

— NOW PLAYABLE
GAME 01 · PLAYABLE
Backgammon
PLAY
GAME 02
COMING SOON
GAME 03
COMING SOON
GAME 04
COMING SOON
A stack of game-AI experiments. Backgammon is playable now — tap the card to take on the engine. More landing soon.
DETAILS & LINKS COMING SOON
IMAGE
IMAGE