Resume
Experience
2025-Present
Expert Contributor, Mathematics
Snorkel AI
Remote
• Authored and validated original, graduate-level problems to test and enhance LLM fluency and reasoning capabilities.
• Contributed to a proprietary dataset of high-rigor challenges used in AI fine-tuning, benchmarking, and reinforcement learning.
• Uncovered failure modes in deductive logic, computation, and mathematical reasoning.
• Advanced the evolution of AI’s mathematical cognition, influencing model fine-tuning strategies and shaping the next generation of AI reasoning benchmarks.
2020-2024
Research Assistant
New Jersey Institute of Technology
Newark, NJ
-
Built and maintained end-to-end ML pipelines for training neural networks on simulation data across Python and MATLAB.
-
Designed and implemented a custom solver for two-phase flow, improving computational efficiency via boundary integral methods.
-
Optimized simulation code for HPC clusters, significantly reducing runtime and scaling performance for large datasets.
-
Developed a high-accuracy numerical method for simulating electroconvective flows using spectral representations.
-
Produced visualizations and simulation videos to support research insights and cross-team collaboration.
-
Collaborated across disciplines, translating mathematical models and computational results into actionable insights.
Projects
1. Developed a fast mesh-free boundary integral method for two phase flow with soluble surfactant. Introduced a fast method for computation of a time convolution integral arising in the Green’s function formulation of the reduced advection diffusion equation, improving operation count to O(N). Provides a novel fast algorithm for advection diffusion. Manuscript submitted to Journal of Computational Physics (2025).
Technical Skills: MATLAB, C, High Performance Computing (HPC)
2. Developed an accurate numerical method for electroconvective flow using a spectrally accurate ultraspherical-Fourier representation. Perform high resolution simulations to obtain training and testing data for convolution neural nets (CNN’s). Build and train CNN’s to predict unclosed terms in an approximate (reduced) model for electroconvective flow.
Technical Skills: MATLAB, HPC, Machine Learning, Python, Pytorch, NumPy, scikit-learn
2019-2024
Teaching Assistant
New Jersey Institute of Technology
Newark, NJ
• Led recitations and coding labs for undergraduate math courses of 20–30 students.
• Designed and graded programming assignments (primarily in MATLAB).
• Provided tutoring and held office hours across multiple undergraduate math and programming courses.
• Proctored and graded exams for lower- and upper-level courses.
Education
2019-2024
New Jersey Institute of Technology
Doctor of Philosophy, Mathematical Sciences
Newark, NJ
Graduated in August 2024. Worked with advisors Michael Siegel and Michael Booty on fluid dynamics problems with a focus on scientific computing and numerical algorithms.
Courses taken include:
Advanced Mathematical Modeling, Methods of Applied Mathematics I-III, Numerical Methods I-II, Optimization, Advanced Scientific Computing, Fast Numerical Algorithms, Stochastic Processes, Machine Learning
​
​
2015-2019
University of Rhode Island
Bachelor of Science, Mathematics
Kingston, RI
Graduated in May 2019 with a minor in Statistics. Played on the women's ice hockey team all four years and was an Academic All American sophomore through senior year.
Courses taken include:
Statistical Methods in Research I-II, Object Oriented Programming, Probability and Statistics, Advanced Calculus and Applications I-II, Numerical Analysis, Data Managing and Programming in SAS, Mathematical Statistics, Multivariate Statistical Learning
​
​
Areas of Expertise
Programming Languages: Python, MATLAB, R, C
​
Tools & Frameworks:
PyTorch, scikit-learn, NumPy, SciPy, pandas, matplotlib, MobaXterm, mex, MATLAB-Python ML pipelines
​
Technical Domains: Machine Learning, Predictive Modeling, Statistical Modeling, Data Analytics, Data Visualization, Data Mining, Deep Learning, ETL, Experimental Design, Scientific Computing, Numerical Methods, Simulation, High-Performance Computing (HPC), Fast Algorithms
Presentations and Publications
S. G. Evans, M. Siegel, J. Tausch, M. R. Booty. "A fast mesh-free boundary integral method for two-phase flow with soluble surfactant." Manuscript submitted for publication, Journal of Computational Physics and arXiv: https://arxiv.org/abs/2506.11282, 2025.
​
S. Evans, A fast mesh-free boundary integral method for two phase ow with soluble surfactant and a study of electroconvective ow, Ph.D. thesis, New Jersey Institute of Technology (2024).