You can view my CV below (updated Jun 2026).
Download CV (PDF)Jeffrey Utley
PhD Candidate, Mathematics — Computational Science
Purdue University, Department of Mathematics | West Lafayette, IN
utleyj@purdue.edu | [Purdue] | [GitHub] | [LinkedIn]
U.S. Citizen
Research Summary
PhD candidate at Purdue University developing statistically validated synthetic data generation algorithms and reproducible scientific software for turbulence-driven imaging and simulation. Three years of DoD-funded research in active collaboration with AFRL and AFIT, with first-author publications in Journal of the Optical Society of America A, Optical Engineering, and SPIE proceedings, and pip-installable Python packages validated on measured datasets. Targeting Research Scientist and R&D roles at national laboratories and in industry, with focus on modeling, estimation, and simulation; computational imaging; signal processing; and scientific software.
Technical Skills
- Methods: algorithm design and analysis; synthetic data generation; inverse problems and parameter estimation; spectral analysis (FFT, PSD, PCA); numerical optimization (first-order methods, Mirror Descent); statistical modeling (Gaussian models, Poisson point processes); Bayesian inference; signal and image processing.
- Optics & Imaging: aero-optics; optical turbulence characterization; atmospheric optics; phase screen generation; wavefront synthesis; boundary layer turbulence; Fourier optics; computational imaging; electro-optical imaging.
- Programming: Python, C/C++, MATLAB, Julia; object-oriented design and modular package architecture.
- Scientific Computing: NumPy, SciPy, pandas, Matplotlib, Jupyter, OpenCV, seaborn.
- Machine Learning: PyTorch, JAX.
- HPC & Systems: Linux/UNIX, bash, SSH, shell scripts, Slurm (Purdue clusters).
- Reproducibility & Tools: Git/GitHub, pip/conda, setuptools/pyproject, Make; reproducible builds, automated testing (pytest), and technical documentation.
Education
Research Experience
- Designed ReVAR, a novel data-driven algorithm producing synthetic data that matches measured statistics from physical experiments with worst-case 4% NRMSE on the temporal power spectrum — a substantial improvement over existing methods.
- Developed BoilingFlow, extending a Fourier-based data synthesis method with novel automated parameter estimation and a spatial-anisotropy extension; validated on measured and simulated datasets (worst-case 12% NRMSE on the temporal power spectrum).
- Built statistical estimation pipelines (FFT/PSD, PCA, Gaussian models) over measured phase screen data spanning >10,000 time-samples per dataset.
- Delivered all results as modular, version-controlled, pip-installable Python packages with reproducible examples and full documentation; ran large-scale parameter sweeps on Purdue HPC clusters via Slurm; produced first-author publications in Journal of the Optical Society of America A, Optical Engineering, and SPIE proceedings.
- Collaborate weekly with DoD program scientists to translate experimental requirements into validated software deliverables.
- Authored four first-author papers (two journal, two conference) in <4 years; presented results through eight conference talks at national conferences.
Summer Research Placements
- Second consecutive AFRL Scholars selection; continues the Purdue–AFRL/AFIT collaboration on synthetic data generation for turbulence-driven imaging and simulation with Prof. Matthew Kemnetz.
- Held the lead computational-science role on the Purdue–AFRL/AFIT collaboration; advanced the BoilingFlow algorithm, delivered weekly virtual technical briefings to AFIT faculty and contributed to a SPIE conference paper deliverable.
- Lead computational-science contributor on the Purdue–AFRL/AFIT collaboration and first AFRL Scholars selection; developed and evaluated the ReVAR algorithm against measured wind tunnel data; delivered an end-of-summer technical briefing to AFRL staff and contributed to a SPIE conference paper.
Research Software
- AOModel — pip-installable Python package implementing the ReVAR algorithm for data-driven synthesis of optical turbulence data. Validated on two wind tunnel datasets with worst-case 4% NRMSE on the temporal power spectrum. Includes thorough documentation and reproducible examples. [GitHub]
- BoilingFlow — pip-installable Python package implementing the BoilingFlow algorithm, extending a Fourier-based data synthesis method with automated parameter estimation and a spatial-anisotropy extension. Validated on two wind tunnel datasets (worst-case 12% NRMSE on the temporal power spectrum). Includes thorough documentation and reproducible examples. [GitHub]
Publications
Refereed Journal Articles
- Utley, J., Buzzard, G., Bouman, C., & Kemnetz, M. ReVAR: A data-driven algorithm for generating aero-optic phase screens. Journal of the Optical Society of America A, 43(7) (in press). [DOI:10.1364/JOSAA.600450]. Preprint: [arXiv:2604.02326].
- Utley, J., Buzzard, G., Bouman, C., & Kemnetz, M. Boiling flow estimation for aero-optic phase screen generation. Optical Engineering (in press). Preprint: [arXiv:2601.12171].
- Lind, J. & Utley, J. (2022). Phase transition for a family of complex-driven Loewner hulls. Involve, a Journal of Mathematics, 15(3), 447–474. [DOI:10.2140/involve.2022.15.447]. Preprint: [arXiv:2106.14940]
Conference Proceedings
- Utley, J., Buzzard, G., Bouman, C., & Kemnetz, M. (2025). Boiling flow parameter estimation from boundary layer data. SPIE Optics + Photonics. [DOI:10.1117/12.3063655]. Preprint: [arXiv:2602.10394].
- Utley, J., Buzzard, G., Bouman, C., & Kemnetz, M. (2024). Data-driven synthetic wavefront generation for boundary layer data. SPIE Optics + Photonics. [DOI:10.1117/12.3027740]. Preprint: [arXiv:2409.04873].
Conference Presentations
- Directed Energy S&T Symposium 2026 — “ReVAR: A data-driven algorithm for generating aero-optic phase screens.”
- Directed Energy Education Workshop 2026 — “ReVAR: A data-driven algorithm for generating aero-optic phase screens.”
- Electronic Imaging 2026 — “ReVAR: A data-driven algorithm for generating aero-optic phase screens.”
- SPIE Optics + Photonics 2025 — “Boiling flow parameter estimation from boundary layer data.”
- Directed Energy S&T Symposium 2025 — “Data-driven synthetic wavefront generation for boundary layer data.”
- Electronic Imaging 2025 — “Synthetic wavefront generation for aero-induced turbulence using boundary layer data.”
- SPIE Optics + Photonics 2024 — “Data-driven synthetic wavefront generation for boundary layer data.”
- Directed Energy S&T Symposium 2024 — “Synthetic wavefront generation for aero-optics correction.”
Honors, Leadership, Service, & Teaching
- Fellowships: Doctoral Research Fellow, Oak Ridge Institute for Science and Education / AFIT (2025).
- Awards: SPIE Student Conference Support Award (2024); Purdue College of Science Travel Award (2024); John H. Barrett Memorial Prize, University of Tennessee (2022).
- Leadership: President, SIAM Student Chapter, Purdue University (2026–Present); Co-organizer, CCAM Lunch Seminar, Purdue Mathematics (2025–Present).
- Service: Reviewer, Optical Engineering (SPIE), 2025–Present (2 manuscripts).
- Teaching: Graduate TA, Purdue University (2022); Instructor, Wolverine Pathways at University of Michigan — math instruction for under-resourced middle and high school students (2022–2023); Undergraduate TA & Math Tutor, University of Tennessee (2020–2022).