CV

• Email: job@camminady.dev
• LinkedIn: camminady
• GitHub: thomascamminady
• Location: Bonn, Germany

Professional Summary

I’m an applied mathematician and data scientist with experience in time series analysis, sensor fusion, and data-driven modeling. At Wahoo Fitness, I develop and deploy algorithms that support training and performance tools used by athletes around the world.

Education

Karlsruhe Institute of Technology, Dr. rer. nat. in Applied Mathematics

Oct 2017 – Jan 2021, Karlsruhe, Germany

• Thesis: Theory, models, and numerical methods for classical and non-classical transport.

RWTH Aachen University, M.Sc. in Computational Engineering Science (CES)

Oct 2013 – Mar 2015, Aachen, Germany

• Selected Courses: Parallel Computing, Computational Differentiation, Uncertainty Quantification, Combinatorial Problems in Scientific Computing.
• Thesis: Theory and application of numerical methods for fractional diffusion equations.

RWTH Aachen University, B.Sc. in CES

Oct 2009 – Sept 2013, Aachen, Germany

• Selected Courses: Computational Fluid Dynamics, Aerodynamics, Calculus of Variations, Partial Differential Equations.
• Thesis: Improvement of the aerodynamic shape optimization by adjoint methods in an MDO process.

Experience

Wahoo Fitness LLC, Data Scientist & Algorithm Developer

Mar 2021 – present, Remote from Germany

• Launched a machine learning–driven training recommendation engine for athletes, enhancing personalized workout strategies.
• Developed robust backend APIs using Python and Pydantic, serving thousands of users via AWS.
• Collaborated with cloud engineers and frontend developers to integrate machine learning models into production systems.
• Applied time series analysis, sensor fusion, uncertainty quantification, and machine learning for rapid prototyping and scalable solutions.
• Analyzed and visualized large-scale, distributed user data using pandas, NumPy, Plotly, and SQL.
• Automated C-code generation for low-level hardware sensors (GNSS, barometer, gyroscope) using Matlab.
• Operated within an agile, fully remote team (US and EU) using Jira and Confluence.

Center for CES & Steinbuch Centre for Computing, Scientific Staff

Apr 2015 – Mar 2021, Aachen & Karlsruhe

• Conducted research in the field of kinetic theory, numerical mathematics, optimization, and machine learning.
• Pioneered reinforcement learning methods to optimize numerical algorithms in CFD, resulting in significant computational improvements.
• Optimized research software on KIT's HPC cluster via parameter studies using OpenMP.
• Served as teaching assistant and substitute lecturer for modules in the mathematics and CES programs.

Festival de Théorie, Summer School on Plasmas

June 2017 – July 2017, Aix-en-Provence, France

• Implemented magnetic field derivatives into a Fortran DG-MHD research code.
• Actively participated in seminars and workshops in the field of plasmas.

Center for CES, Student Research Assistant

Jan 2010 – Mar 2015, Aachen, Germany

• Assisted with teaching duties for a variety of mathematics and computer science modules.
• Developed GPU-based Monte Carlo simulations using CUDA.

EADS Cassidian, Internship with Bachelor's Thesis

Oct 2012 – Apr 2013, Manching, Germany

• Performed numerical simulations with in-house tools and the adjoint code of the German Aerospace Center (DLR).
• Automated UAV airfoil shape optimization using mesh adjoints.

Projects

• LeTourDataSet

• alltime athletics python

• traffic-balve

• dicopal

Scientific Outreach

Collaborator in the Computational and Mathematical Modeling Program (KIT University). Developed educational programs that demonstrate the importance of mathematical modeling and machine learning for real-world applications to high-school and entry-level university students. Authored publications in mathematical didactics aimed at integrating mathematical modeling into the German Abitur curriculum.

Skills

• Python: 8 years of experience: Expert in object-oriented and type-annotated Python, employing advanced design paradigms for robust development and testing. Experience with NumPy, SciPy, pandas, matplotlib, and scikit-learn.
• MATLAB: 6 years of experience: Development of scientific simulation tools and visualizations. Includes working with MuPAD, Simulink, and Optimization Toolbox.
• Data Visualization: 10+ years of experience: Expert in creating interactive data visualizations and dashboards using Altair, Plotly, Matplotlib, D3.js, and Observable Plot to present complex data in an accessible way. This includes publication-ready visualizations, interactive visualizations for explorative analysis, and development of deployable dashboards.
• Software Development: 10+ years of experience: Version control (Git, Github), CI/CD, modern testing frameworks, proficiency with UNIX systems, AWS (Lambda), Jira, and Confluence.

Publications

Mathematische Grundlagen der Künstlichen Intelligenz im Schulunterricht

• Sarah Schönbrodt, Thomas Camminady, Martin Frank
• Mathematische Semesterberichte 69 (1), 73-101, (2022)

Theory, models, and numerical methods for classical and non-classical transport

• Thomas Camminady
• Dissertation (2021)

Ray Effect Mitigation for the Discrete Ordinates Method Using Artificial Scattering

• Martin Frank, Jonas Kusch, Thomas Camminady, Cory D. Hauck
• Nuclear Science and Engineering, Vol. 194, No. 11, pp. 971–988 (2020)

Vorschlag für eine Abiturprüfungsaufgabe mit authentischem und relevantem Realitätsbezug

• Sube, Maike, Thomas Camminady, Martin Frank, Roeckerath, Christina
• Modellierungskompetenzen – Diagnose und Bewertung, Springer Berlin Heidelberg, pp. 153–187 (2020)

Ray effect mitigation for the discrete ordinates method through quadrature rotation

• Thomas Camminady, Martin Frank, Kerstin Küpper, Jonas Kusch
• Journal of Computational Physics, Vol. 382, pp. 105–123 (2019)

Highly uniform quadrature sets for the discrete ordinates method

• Thomas Camminady, Martin Frank, Jonas Kusch
• Proceedings of the International Conference on Mathematics and Computational Methods Applied to Nuclear Science and Engineering, pp. 25–29, (2019)

A spectral Galerkin method for the fractional order diffusion and wave equation

• Thomas Camminady, Martin Frank
• International Journal of Advances in Engineering Sciences and Applied Mathematics, Vol. 10, No. 1, pp. 90–104 (2018)

A new high-order fluid solver for tokamak edge plasma transport simulations based on a magnetic-field independent discretization

• Giorgiani, G., Thomas Camminady, Bufferand, H., Ciraolo, G., Ghendrih, P., Guillard, H., Heumann, H., Nkonga, B., Schwander, F., Serre, E., Tamain, P.
• Contributions to Plasma Physics, Vol. 58, Nos. 6–8, pp. 688–695 (2018)

Nonclassical particle transport in heterogeneous materials

• Thomas Camminady, Martin Frank, Edward W. Larsen
• Proceedings of the International Conference on Mathematics & Computational Methods Applied to Nuclear Science & Engineering (2017)

The equivalence of forward and backward nonclassical particle transport theories

• Edward W. Larsen, Martin Frank, Thomas Camminady
• Proceedings of the International Conference on Mathematics & Computational Methods Applied to Nuclear Science & Engineering (2017)

Theory and application of numerical methods for fractional diffusion equations

• Thomas Camminady
• Master's Thesis (2015)

Improvement of the aerodynamic shape optimization by adjoint methods in an MDO process

• Thomas Camminady
• Bachelor's Thesis (2013)