Summary
Andrew Mcdonald is a quantitative developer with 14 years of experience combining academic rigor from a Drexel CS Ph.D. program with production-focused trading and risk systems. He builds end-to-end, low-latency algorithmic platforms and backtesting frameworks—leveraging TimescaleDB/Postgres, TensorFlow (including TensorFlow Probability), and IB-insync—to optimize trade-level expected value while controlling downside volatility. At Interactive Brokers he applies quantitative risk engineering in a production environment, and previously developed bespoke ML-driven strategies spanning Bayesian nets, deep RL, SVMs, and Markov regime models. His research pedigree includes novel neural architectures (sparse super-regular networks), graph-conv + wavelet forecasting for environmental time series, and a biologically-inspired parallel neuron simulator, reflecting a strong ability to move ideas from theory to robust implementations. An active contributor to open-source tooling, he has also maintained core analysis code for the Anonymouth authorship-anonymization project, highlighting a cross-domain fluency in ML, NLP, and systems engineering.
14 years of coding experience
12 years of employment as a software developer
Bachelor of Science (B.S.), Computer Engineering, Cum Laude, Bachelor of Science (B.S.), Computer Engineering, Cum Laude at Drexel University
java, python, c++, c, php, javascript, common lisp, mysql, bash