Dereck Padden

Sr ML Engineer.

United States

Summary

ML Engineer with solid experience in implementing and prototype algorithms looking for solid remote opportunities. Skilled with Python and multithreaded C++.

Favorite Python Packages:

scipy, skearn, pandas, pytorch, boto3

Experience

Neurable, Inc. – Boston, MA
Senior Machine Learning Engineer, October 2016 – Present
-Contributed to the development of the Neurable C++ Brain Computer Interface API on Linux
-Wrote multi-threaded C++ code to enable real-time data processing of EEG data
-Developed Python asynchronous test platform for the prototyping of EEG machine learning and
signal processing algorithms
-Prototyped signal processing and machine learning algorithms in Python
-Designed and implemented the above software pipelines for the Neurable C++ API
-Experimented with ML methods based on Symmetric Positive Definite (SPD) matrices and
Riemannian Geometry
-Implemented Unit, Integration, and Performance tests using Google Test framework
-Utilized cloud computing (Docker, AWS) to perform batch processing of performance metrics
-As needed, maintained CMake build scripts
-Wrote visualization scripts for heuristic data quality control assessment to aid with data collection
-Participated in study design and architectural planning discussions
-Engaged with product team to help translate customer requests into actionable tasks
Innerscope Research/Nielsen Consumer Neuroscience – Boston, MA
R&D Specialist - Programmer ― Operations and Analytics, February 2015 – July 2016
-Provided analytical software support and algorithm expertise for Neuromarketing analysis
-Developed internal software for the analysis of Biometric data (Galvanic Skin Response, Heart Rate)
-Refactored legacy MATLAB code to improve maintainability and modularity
-Ported open-source MATLAB analysis library into Python (SciPy stack)
-Wrote multiple Python scripts and function libraries to automate analysis tasks (SciPy, Pandas)
-Designed prototype algorithm for custom Skin Conductance analysis tool for NBC Universal
-Participated in metrics validation study using sales data provided by Nielsen Catalina
-Designed Support Vector Machine (SVM) based estimator (SciKit-Learn) to rapidly assess Quality
Control of input Biometric signals using spectral features
-Developed GUI using PyQT and SciPy stack to perform Quality Control and train SVM classifierW2 Contractor at Apex Systems, Inc. in support of MIT Lincoln Laboratory – Lexington, MA
Research Engineer - Developer ― SONAR Laboratory, June 2011 – February 2015
-Evaluated algorithm performance for Adaptive SONAR Beamforming in a MATLAB environment
-Contributed to the development and implementation of new approaches and algorithms
-Modified existing algorithms for use with different systems
-Performed scenario analysis
-Converted Kalman Filter tracking algorithms written in MATLAB into C++
-Wrote/modified binary data reading software in MATLAB and C++
-Modified C++ code to augment TCP/IP communication capabilities between components within an
autonomous system
-Using existing C++ TCP/IP drivers, created external packet monitoring interface for concurrent use
with real-time updated MATLAB display
-Implemented polyphase downsampling and upsampling filters in C++ for use in real-time system
Raytheon IDS ― Woburn, Massachusetts
Senior Systems Engineer - Analyst ― Missile Defense Radar Systems, Oct 2008 – May 2011
-Performed task lead and technical duties for trade studies to improve RADAR object classification
-Implemented various algorithms in a MATLAB test environment to determine performance benefit
-Wrote classification, feature selection, and training scripts
-Employed information theory and clustering methodology during exploratory analysis
-Configured and created scenarios for RADAR simulations
-Participated in interchange meetings with customers to discuss architecture, progress, and results
Raytheon NCS ― Marlborough, Massachusetts
Electrical Engineer II ― Test Systems, June 2007 – October 2008
-Designed and integrated test station electrical hardware
-Designed/composed test station software in Visual C, LabView, LabView FPGA
-Debugged Digital I/O cards and Visual Basic software
-Wrote/Revised test requirement specification documents
-Cooperated with a team of systems, electrical, and mechanical engineers to achieve
designs which met customer demands
Harris Corporation RF Communications ― Rochester, New York
Firmware Test Co-op, Summer 2005
-Utilized Visual Basic scripts and lab hardware to test military radio firmware functionality
-Conducted manual field testing of radios
-Analyzed test failures in conjunction with engineers in order to expose firmware bugs