Resume

Link to 1-page PDF version

EDUCATION

TECHNICAL SKILLS and CLASSES

Skills: Python, PyTorch, Bash, C++, MPI, LAMMPS, Dask, MongoDB

Classes: Deep Learning, Applied Machine Learning, Atomic Scale Simulations, Parallel Programming, Numerical Analysis, Mathematical Methods in Physics, Advanced Applied Mathematics, Object-Oriented Programming and Design

RESEARCH and PROFESSIONAL EXPERIENCE

UIUC: Ph.D. Researcher -- Urbana, IL

Aug 2017 - Present

• Designed a machine learning model for describing atomic interactions (interatomic potential) using spline-based atomic environment descriptors with speeds up to 10x faster than models with similar accuracy.

• Developed a novel method for constructing universal interatomic potentials by using generative models (variational autoencoders) for building transferable and interpretable latent spaces.

• Wrote parallel, scalable fitting software from scratch using Python, PyTorch, and Dask; wrote software patches in C++ for existing production-level molecular simulation software (LAMMPS).

ColabFit (University of Minnesota): Ph.D. Research Intern -- remote

Sep 2021 - Present

• Contributed to research community and enabled large-scale analysis of materials data by leading a community-wide effort to design and construct an open-source database of electronic structure calculations.

• Established a standard for storing first-principles simulation data in an efficient and queryable format, and constructed Python and web APIs to a Mongo database of user-contributed datasets.

• Directed a team of researchers in design and development of software framework, and interfaced with users in the field to guide research direction and collect feedback.

UofA: Undergraduate Research Assistant -- Tucson, AZ

Sep 2016 - Aug 2017

• Analyzed mechanical and thermal properties of self-assembled buckminsterfullerene (C60) structures using computational methods.

• Designed, executed, and interpreted molecular dynamics simulations of bulk C60 molecular solids.

• Applied basic statistical techniques to analyze simulation results, and wrote software tools in Python to aid in supercell construction.

Sandia National Laboratories: Student Intern (year-round) -- Albuquerque, NM

June 2016 - March 2017

• Developed object recognition software for addressing unique challenges associated with X-ray images by experimenting with various feature detector and descriptor algorithms in Python and MATLAB.

• Summarized and presented results at the 2016 SPIE Optics + Photonics conference.

UofA: Undergraduate Research Assistant -- Tucson, AZ

Jan 2015 - Aug 2016

• Contributed to NASA research project analyzing directional solidification of aluminum - 7% silicon alloys on the International Space Station to study the effects of convection on macrosegregation and dendrite arm spacing.

• Performed fundamental metallography tasks of Al-Si alloys, including sample polishing and preparation, imaging, image analysis. Wrote basic image analysis software in Python for quantifying dendrite growth.

UofA: Undergraduate Teaching Assistant -- Tucson, AZ

Jan 2015 - March 2016

• Helped prepare incoming freshmen for college-level algebra by conducting a weekly online classroom and facilitating in-person office hourse to provide math tutoring and academic advisement.

• Trained new teaching assistants by coordinating and leading training sessions about teaching best-practices and communication skills.

PUBLICATIONS

1. Joshua A. Vita, Dallas R. Trinkle, "Exploring the necessary complexity of interatomic potentials", Computational Materials Science, Volume 200 (2021), 110752, https://doi.org/10.1016/j.commatsci.2021.110752.

2. A. Alsayoud et al., "Atomistic insights into the effect of polymerization on the thermophysical properties of 2-D C60 molecular solids", Carbon (2018): DOI: 10.1016/j.carbon.2018.01.044

3. J. Vita et al., "Hybrid object detection system for x-ray radiographs", SPIE Optical Engineering + Applications Proceedings, Volume 9969, Radiation Detectors: Systems and Applications XVII; 996909 (2016): DOI: https://doi.org/10.1117/12.2237213

4. A. Wantuch et al., "Exploration of available feature detection and identification systems and their performance on radiographs", SPIE Optical Engineering + Applications Proceedings, Volume 9969, Radiation Detectors: Systems and Applications XVII; 996907 (2016): DOI: https://doi.org/10.1117/12.2237211

PRESENTATIONS

• "Interpretability in Deep Learning Models for Atomic-Scale Simulations", Lawrence Livermore National Lab Data Science Institute, virtual, July 2022 (link)

• "Interpretability in Deep Learning Models for Atomic-Scale Simulations", Los Alamos National Lab, Los Alamos, NM, July 2022

• "Exploring the Necessary Complexity of Interatomic Potentials", 2022 Spring Materials Research Society Meeting, Honolulu, HI, May 2022

• "ColabFit: Collaborative Development of Data-Driven Interatomic Potentials", APS March Meeting 2022, Chicago, IL, March 2022

MEMBERSHIPS, ACTIVITIES, and HONORS

• UIUC NSF DIGI-MAT fellow (link)

  • Fellow and leadership board member -- NSF-funded Ph.D. training program for promoting data science in materials research; served as student/professor liaison to aid in program development, course design, and student on-boarding

• 2018 African School on Electronic Structure: Methods and Applications (ASESMA)

  • Student instructor in two-week workshop in Ethiopia to promote research using DFT at African universities. Led workshops and discussions to teach students how to design and run molecular dynamics simulations

• The Hacker Within

  • Computational science student club for promoting collaboration and open discussion of various topics including programming, machine learning, cybersecurity, and data visualization. Software Carpentry Leader: conducted training sessions teaching Python fundamentals, introductions to Github, and code optimization.