Marissa D'Alonzo

MathWorks

Responsible for developing online, self-paced learning courses for online platforms such as Coursera and EdX

Draper Laboratory

Implemented a parametric Gaussian Mixture Model neural network to quantify the uncertainty associated with a normalized-cross correlation match during terrain relative navigation for the DARPA Enabling Confidence program. Prepared the input data using numpy and trained the model using PyTorch. This network was fully integrated with the navigation system using ONNX and the OpenCV DNN library and showed improvements in maximum navigation error and reduced reliance on GPS.
Created a diverse black box simulation generation framework to thoroughly study the potential for adversarial machine learning to find failures in simulation.
Managed project direction, budget, and customer communication as the technical director and subject matter expert on a SBIR contract to develop a hazard detection system for landing on irregular surfaces.
Served as the subject matter expert on lidar-based hazard detection algorithm for the Dynetic's Human Landing System team. Implemented algorithm improvements in C++, conducted simulation and testing, and worked with systems engineering to quantify the requirements of the system.
Leverage Kconfig, bash, CMake and GNU make to extend the capability of an internal package management tool used company-wide to handle new packages and operating systems.

Worked in conjuction with Memorial Sloan Kettering Cancer center to develop weakly supervised methods for the interpretation of RCM mosaics for skin cancer identification

Implement three crater matching algorithms into existing C++ codebase - two from scratch, one from an exisiting algorithm
Develop crater projection function in C++
Update graphical user interface to reflect new matching algorithms

Spring - Summer 2019 Draper

Conducted experiments profiling power capping on an Amazon Web Services supercomputing node
Analyzed data collecting using Python to profile applications and discover connections between variables

Fall - Spring 2019 Northeastern

Conducted research and analysis on the different forms of terrain data available to the laboratory
Condensed library volumes into two page informational packets for future reference
Developed and conducted analyses and tests to measure the completeness and accuracy of data types
Modified existing libraries to accomodate different file types

Spring - Summer 2018 MITLL

Created and maintained databases in Microsoft Access of students enrolled in programs
Assisted students participating in the Swarmathon Challenge
Developed and analyzed swarm search algorithms

Researched, designed, and simulated a variety of circuits to determine the safest and most cost-effective
Modified PCBs and hand-made cables, culminating in a working prototype of a spinal table
Designed and debugged test setups using Quantum Leaps and custom PCBs
Monitored and analyzed data from test setups

Spring-Summer 2017 Allen

Work Experience