Sam Duffy

Skills

Languages: Python, Q, SQL, KDB, MATLAB

Data & Tools: Tableau, Azure Blob Storage, Bloomberg Data License

Platforms & Methods: Linux, Kubernetes, Optimization, Statistics, Applied Machine Learning

MFS Investment Management — Boston, MA

Quantitative Trading Specialist — Systematic Execution Team Apr 2023 – Present

Senior Software Engineer — Systematic Execution Team Dec 2021 – Apr 2023

Spearheaded a custom implementation of KDB Insights, using Python, Q, and Kubernetes to process internal order models and TAQ data of 15,000 securities into KDB.
Collaborated with portfolio managers and traders to provide actionable insights on the trading cost impact of order decisions, broker toolkits, and global security selection.
Leveraged statistics to create a multi-faceted Algo Wheel optimizer to reduce the trading cost of the systematic order flow.
Derived a closed-form solution for the Algo Wheel true-up process, reducing monthly allocation error to below the 5th percentile of the previous solution when validated with Monte-Carlo simulations.
Developed an internal Impact Cost Model for equities.
Created complex Tableau dashboards to give equity traders insight into broker algo performance.
Developed internal Python libraries to enable rapid prototyping and development of various model pipelines with optional features like bootstrapping, bagging, and stratified sampling.
Developed Python libraries to extract EMS execution and order data via API endpoints.
Created a custom job scheduler to automate all daily, weekly, and monthly jobs for the TDA Team.
Used KNN clustering and stratified resampling to compare broker performance across diverse order flows.

Raytheon Missile & Defense — Tewksbury, MA

Systems Engineer — PATRIOT Surveillance Radar Jun 2020 – Dec 2021

Developed a suite of tools for rapid data extraction, display, and analysis using OOP techniques.
Completed root-cause analysis of defects spanning all parts of the PATRIOT Surveillance systems.
Led the test team of an internal Radar & Missile Simulation project, designing and executing the functional requirement test plans.
Automated the annual verification process of complex functional requirements, reducing weeks of analysis to under a minute.
Led a team of engineers to integrate several custom machine learning models to detect anomalous flight patterns for air traffic controllers.
Used TensorFlow and Keras to develop a custom TadGAN machine learning model and training functions for each of its networks.

The Applied Research Lab — State College, PA

Distinguished Undergraduate Researcher May 2018 – May 2020

Completed a thesis on applied swarm intelligence, documenting multi-modular neural network architectures, meta-heuristic optimization algorithms, and multi-agent systems.
Developed an original algorithm to model the stalk-and-ambush hunting behavior of lions for multi-predator, single-prey scenarios.

The Pennsylvania State University Spring 2020

B.S. in Mechanical Engineering — University Park, PA