About Me

I am a recent Physics PhD making a change of career into Machine Learning and Artificial Intelligence. I have a strong quantitative background, with specific expertise in high performance computing, numerical methods and data analysis.

Education:

• Ph.D. Physics (University of Colorado, Boulder, 2017): “Perturbative and Ab-initio calculations of the electrical susceptibilities of atoms”
• M.S. Physics (University of Colorado, Boulder, 2013).
• B.S. Physics (University of Washington, 2010).

Professional Experience:

2011 - 2017: Research Assistant in JILA at CU Boulder

• Programming: Developed a non-trivial code to solve the time dependent Schr"odinger equation (TDSE) in a field-free energy basis. The code was developed using PETSc and Boost MPI in C++14, with an approximate size of 10kloc. https://github.com/spott/ebss
• Data Analysis: Data analysis package for the above TDSE solver output. Developed using pandas, numpy, scipy and matplotlib. Package available at https://github.com/spott/python_da_lib/

2015: Data Science Intern for Cognilytics at Centurylink

• Data Science: Preliminary analysis of anomalous network traffic for network security applications using raw packet captures with support vector machines, recurrent neural nets and random forests.

2010 - 2017: Teaching Assistant

• Communication: Taught a wide variety of introductory physics classes. Demonstrated a strong ability to take complex mathematical concepts and explain them intuitively.

Skills

• Python Data analysis using NumPy, SciPy, pandas, matplotlib.
• Skilled in Python and C++14

Relevant repositories

• webpage-classifier - A project to classify webpages by topic. Uses reddit posts as a labeled dataset. Utilizes Docker, SQL, and async technologies in python. Latent Dirichlet Analysis via gensim is used to perform
• petsc-cpp - A C++ wrapper around the Sparse Linear Algebra Library PETSc.
• ebss - A solver for the time dependent Schrodinger equation of an atom interacting with a strong laser field using the field-free energy basis. This is developed using PETSc and Boost MPI in C++14. It leverages MPI in order to efficiently run on large supercomputers.
• da_lib - A set of tools for analysis of data from the above EBSS.

Publications

1. A. Spott, A. Jaron-Becker, and A. Becker: Ab initio and perturbative calculations of the electric susceptibility of atomic hydrogen
2. A. Spott, A. Becker, and A. Jaron-Becker: Transition from perturbative to nonperturbative interaction in low-order-harmonic generation
3. A. Spott, A. Jaron-Becker, and A. Becker: Time-dependent susceptibility of helium atom in intense laser pulses} (in review.)