Data scientist and software engineer with 12 years of experience in computational research and open-source infrastructure for neuroscience. Curated 256 TB of datasets on NIH archives, built 12+ data pipelines for research labs, and currently maintain petabyte-scale cloud archives at Dartmouth. Trained researchers from dozens of universities through conference workshops and hackathons. Deep toolkit in Python, AWS, DevOps, and cloud-optimized data formats (HDF5, Zarr), with a Ph.D. foundation in time series analysis, nonlinear dynamics, and machine learning.
Relevant coursework: computational neuroscience, machine learning, Bayesian statistics, network science, stochastic analysis, time series analysis, partial differential equations, nonlinear dynamics
Relevant coursework: calculus, linear algebra, probability, statistics, discrete & combinatorics, numerical analysis, real analysis, mathematical physics, algorithms & data structures, cryptography, machine learning, symbolic logic, cognitive psychology, neurophysiology, advanced neuroscience, philosophy of mind, philosophy of science
Maintaining and enhancing several NIH data archives that store neuroscientific data by standardizing dataset organization across multiple modalities and providing usage analytics, including geolocation, through parsing terabytes of raw S3 log files.
Assisting with user support tickets and feature requests to ensure smooth operation of data submission and retrieval processes.
Developing open source software for use in neurophysiology research with an emphasis on achieving reproducibility of scientific findings.
Researching options for stable, cost-effective, long-term backup of petabyte-scale data storage. Working with a range of institutions and services to provide solutions and cost estimates to that effect.
Providing software, consultation, and computing support for the Brain Behavior Quantification and Synchronization consortium (BBQS).
Developed and maintained 5 repositories of open-source software by ensuring proper functionality of automated testing suites, documentation, tutorials, and demos.
Created and maintained 12+ data processing pipelines for neuroscience labs, allowing their data to flow from acquisition to sharing in a seamless fashion.
Curated a total of 256 TB of high-value datasets to NIH data archives on behalf of various research groups.
Managed the company's cloud resources (Amazon Web Services; AWS), including storage, compute resources, and identity access management (IAM).
Handled user interactions within the research community by offering technical support and resolving issues or feature requests in a timely manner.
Facilitated user education across various platforms by running sessions at multiple conferences and workshops, increasing user adoption and effective system utilization.
All software related in some way to the facilitation of terabyte-scale data management, analysis, and visualization for the field of neurophysiology.
Reconciled the computational properties of biologically realistic neural networks with artificial machine learning models (such as those used across computer vision) through complex mathematical theory and stochastic biophysical simulations, with results communicated through 3 journal publications and a presentation at the high-impact COSYNE conference (top 4% of abstracts accepted).
Collaborated with several top experimentalists in neuroscience as a trainee in the NeuroNex program, which focused on understanding how neural function emerges from underlying structure.
Ran 4 tutorial sections for 112 students, aiding their study of course concepts to achieve a 96% satisfaction rate.
Graded homework assignments and exams. Filled in main lectures for the professor as needed.
Analyzed geological data from water samples in conjunction with the Indiana Department of Environmental Management to issue compliance permits for Indianapolis regulation standards.
Reviewed and corrected over 300 pages of "An Idiot's Guide to Algebra II".
Explored statistical effects of algorithmic curation (the use of automated filtering mechanisms in the delivery and display of information) by measuring properties of simulated models of social networks, with results presented at the MIDSURE conference.
Examined information diffusion through large-scale simulations of G-protein signaling mechanisms using a high-performance super-computing cluster (HPC), then presented results at an undergraduate symposium.