Headshot of Michael Kline

Michael Kline

Data Scientist/Modeler II

michael@mkline.co LinkedIn Download résumé (PDF)

Results-driven data scientist and computational modeler who leads cross-functional teams through the full data lifecycle — from GPU-accelerated simulation and probabilistic risk modeling to executive-level reporting that shapes major investment decisions. Physics background with published research from Oak Ridge National Laboratory.

$100M+
in investment decisions informed by my analyses
$1.3M+
annual revenue across the project portfolio I manage
15+
data scientists & modelers led on one cross-functional team
40×
speedup delivered on mission-critical client code

Experience

Battelle Memorial Institute
Columbus, OH · Nov 2022 – Present

  1. Data Scientist/Modeler II

    Jan 2024 – Present Current
    • Manage and direct a portfolio of multi-million-dollar technical initiatives ($1.3M+ annual revenue), driving continuous operational improvement
    • Lead a cross-functional team of 15+ data scientists and modelers developing a robust, large-scale probabilistic risk assessment to guide strategic investments
    • Instituted best practices for code governance and documentation (CI/CD principles), significantly reducing technical debt and improving maintainability across the modeling group
    • Synthesize complex analytical findings into high-impact executive presentations, directly influencing investments valued over $100 million

  2. Data Scientist/Modeler I

    Nov 2022 – Jan 2024
    • Engineered performance improvements for mission-critical, client-facing code, achieving a 40× speedup in processing time while maintaining rapid turnaround
    • Developed a custom file format optimized for departmental workflows, eliminating manual data-preparation steps and reducing latency for cross-functional teams
    • Translated complex statistical models into clear visualizations and executive summaries for technical and non-technical stakeholders, enabling faster decisions

Skills

Languages & computing

  • Julia
  • Python (NumPy, SciPy, Pandas)
  • CUDA / GPU computing
  • C++
  • C# / .NET
  • SQL
  • MATLAB

Modeling & analysis

  • Probabilistic risk assessment
  • Physics-based simulation
  • Statistical modeling
  • Data visualization
  • COMSOL Multiphysics

Data & DevOps

  • PostgreSQL
  • Custom file-format engineering
  • Git & GitHub Enterprise
  • Azure DevOps
  • Docker

Leadership & delivery

  • Technical team leadership
  • Project portfolio management
  • Executive communication
  • CI/CD & code governance
  • Performance optimization

Education & Certifications

The Ohio State University
Aug 2021 – Nov 2022

M.S., Physics

Transylvania University
Sep 2018 – Dec 2020

B.A., Physics & Mathematics

  • Summa Cum Laude · 4.0 GPA
  • Advisor: Stephen Johnson
  • Google Data Analytics Professional Certificate
    Coursera
    Verify

Publications

Improved limits on n → n′ transformation from the Spallation Neutron Source

Physical Review D 110, 072022 Oct 30, 2024
Abstract

Conversions between neutrons n and dark matter candidate sterile neutrons n′ have been proposed as a mechanism for baryon number violation. In the case that there is a small mass difference Δm between the n and the n′ states, oscillations can be induced by compensating for Δm with a magnetic field. A search for such neutron oscillations was performed at the Spallation Neutron Source by looking for anomalous neutron transmission through a strongly absorbing cadmium wafer inside of a 6.6 T magnet. The approach described here saw no regenerated neutrons above background, which provides an improved limit for neutron–sterile neutron transformations for a range of Δm between 0.1 and 1000 neV.

PDF DOI

Experimental Search for Neutron to Mirror Neutron Oscillations as an Explanation of the Neutron Lifetime Anomaly

Physical Review Letters 128, 212503 May 27, 2022
Abstract

An unexplained >4σ discrepancy persists between “beam” and “bottle” measurements of the neutron lifetime. A new model proposed that conversions of neutrons n into mirror neutrons n′, part of a dark mirror sector, can increase the apparent neutron lifetime by 1% via a small mass splitting Δm between n and n′ inside the 4.6 T magnetic field of the National Institute of Standards and Technology Beam Lifetime experiment. A search for neutron conversions in a 6.6 T magnetic field was performed at the Spallation Neutron Source which excludes this explanation for the neutron lifetime discrepancy.

PDF DOI

Research Experience

GPU Simulation & Data Analysis — Mirror Neutron Search

Jan 2021 – Aug 2021
Oak Ridge National Laboratory (SULI) Advisor: Leah Broussard

Built the simulation and performed the data analysis for the search for mirror neutrons. Used CUDA and Julia to compute detection probabilities across tens of millions of points in the parameter space — work that contributed to publications in Physical Review Letters (2022) and Physical Review D (2024).

  • Julia
  • CUDA
  • Large-scale parameter sweeps
PRL paper PRD paper Presentation

Neutron Spin-Tracking Simulation — nEDM@SNS

Jun 2020 – Aug 2020
Oak Ridge National Laboratory (SULI) Advisor: Leah Broussard

Created a GPU simulation tracking neutron spin in the measurement cell to study systematic effects that could produce a false EDM measurement. Presented at the 2020 APS Division of Nuclear Physics Fall Meeting.

  • Julia
  • CUDA
  • Systematic-error analysis
Report Poster Presentation

Physics Education Research — Master's Thesis

Aug 2021 – Nov 2022
The Ohio State University Advisor: Andrew Heckler

Analyzed student response data to measure how answer format affects learning and assessment of vector addition in introductory physics; found free-response formats offered clear benefits for both.

  • Statistical analysis
  • Education data
Master's Paper Presentation

Photothermal Deflection Spectroscopy Simulation

May 2019 – Dec 2020
Transylvania University Advisor: Stephen Johnson

Modeled heat deposition and temperature dispersion in COMSOL Multiphysics, then ray-traced light through the cuvette in Julia to predict the detector signal. Presented at COMSOL Conference 2019 Boston.

  • COMSOL
  • Julia
  • Ray tracing
Poster

Electromagnetic Coil Design — nEDM Experiment

Jan 2019 – Aug 2019
University of Kentucky Advisor: Christopher Crawford

Solved for and iteratively refined coil windings using equipotential contours of the magnetic scalar potential along the coil boundary. Presented at COMSOL Conference 2019 Boston.

  • COMSOL
  • MATLAB
  • Iterative optimization
Code Poster

Conference Talks & Posters

  • AAPT Summer Meeting
    Talk · Jul 2022
    Slides
  • APS Division of Nuclear Physics Fall Meeting
    Talk · Oct 2021
    Slides
  • APS Division of Nuclear Physics Fall Meeting
    Talk · Oct 2020
    Slides
  • COMSOL Conference Boston
    Two posters · Oct 2019
    Transylvania Kentucky

Let's talk

Interested in working together? I'd be glad to hear from you.

michael@mkline.co LinkedIn Résumé (PDF)