John Edmiston Email & Phone Number

Houston, Tx, Us

Developed solutions to satellite mission planning, including tasking and comms, as part of a SaaS product.

Cambridge, Massachusetts, Us

- Deployed video-data based PyTorch CNN to improve pipeline efficiency using AWS Batch- Developed POC and architecture test suite for cloud based real time inference application

Hoboken, Nj, Us

Developed POCs for incubator arm, centered on a nutrition product.

Emeryville, California, Us

Developed in house MES for task scheduling and recording tool (Flask application) to record data from production operators. Scheduling included accounting for physical location, time, and worker shift information. System centralized the labeling and storage of multiple sources of data including photos, multiple choice questions, and democratized data streams across functions. Made MVP and successfully cut over to production architecture. Utilized GCP Pub/Sub for IoT edge communications with several hardware devices. Developed logging system for multiple Raspberry Pi sensor devices, including automatic setup from cold start and rolling out software updates to all devices. Used GCP methods like PubSub, Cloud functions, Celery to perform tasks asynchronously and keep application responsive.Digitization efforts led to 30% improvement in per-worker productivity, and 225% increase in total managed capacity.Hardware/software IOT system designer and builder, a multi-measurement computer vision based system, sensor logging architecture, machine learning on sheet quality using GCP Pub/Sub, Celery, Javascript/D3.js, Docker

Redwood City, Ca, Us

Developed software to combine/munge data streams (lab records, EHR) from multiple clinical providers for candidate selection.

San Francisco, California, Us

Aggregation and reporting of usage data for business performance. Developed pipelines and munging in Python/R. Analyzed performance of experiments via multilevel hierarchical models. Efforts to automate reporting functions resulted in serving as the only data hire during the first rapid growth phase (from Series A-B).

Livermore, Ca, Us

- Managing FEM simulations on HPC Linux clusters with Python/Bash scripting- Developed custom post processing scripts with VTK/Visit/Python/MPI- Meshing complex CAD geometries with CUBIT, including Python scripting

Berkeley, Ca, Us

Developing modeling and simulation capabilities for coupled flow and deformation in porous materials. These include finite element methods for geomechanics using ROCMECH, with porous flow computation using the TOUGH family of codes. I am also working on a geoperidynamics implentation for fracture propagation problems. Coding environment: Fortran77, Fortran90, Fortran2003, C, C++, PythonLibraries: PETScFor a commonly demonstrated example for basic peridynamics see a simulation video I uploaded at https://youtu.be/Arh4YuTbj7IFor multiphysics see my implementation of the Element Free Galerkin for the Terzaghi consolidation problem. https://www.youtube.com/watch?v=s5OCsctwRNg&feature=youtu.be

Research and develop numerical meshfree methods for large deformation, high velocity impact simulation. Developed novel stabilization scheme for MLSPH method. Implemented parallelization using OpenMP and MPI. Code development: C++ for simulations, post processing using Python/SciPy/NumPy/Matplotlib.

Livermore, Ca, Us

Developed uncertainty quantification approach for X-ray diffraction analysis of individual grains in a polycrystal. Constructed a model and alternative constitutive framework for single crystal plasticity, with calibration to data. Code development for X-ray diffraction data analysis: Python/NumPy/SciPy/C++ hybrid code via Boost.Python.See the attachments below for draft publications from this time period.

Doctor Of Philosophy (Ph.D.), Mechanical Engineering

Master Of Science (M.S.), Mechanical Engineering

Bachelor Of Science (B.S.), Mechanical Engineering