Software Engineer, Robotics — Anduril
I design and build software and autonomy for robots deployed around the world.
Software Engineer — Robotics, Machine Learning & Computer Vision
I have 4 years of industry experience building robotics software for industrial and military applications, and have also dabbled in bioinformatics research. I aim to develop intelligent systems that improve the quality of life and safety of humans around the world.
I'm passionate about the development of robotics & ai. I am most engaged when working across the entire design and development process, from research to deployment. When I'm not building, you'll find me at the gym, with friends, or recovering at the sauna.
languages: C++, Python, (~Rust) · theory: CS, ML, Vision · tools: Linux, tmux, Neovim
I design and build software and autonomy for robots deployed around the world.
Designed and built software and algorithms for military and industrial applications. I was responsible for systems across the stack, including backend microservices, computer vision pipelines (object detection & tracking), user interfaces, and system integration. I led development on various projects including an R&D object tracking system that involved ML model creation and a custom tracking algorithm based on kalman filters. I worked closely with cross-functional teams and customers to solidify requirements and guide development.
Led technical contribution on a paper that achieved state-of-the-art results in predicting progression of oral cancer. Used current literature to guide design and development of models. Trained, finetuned and ran inference on custom PyTorch models. Analyzed model results and data characteristics to pinpoint model weaknesses and improve performance ~0.20 F1. Tracked experimental outcomes and metrics, produced publication-level manuscript visualizations, and contributed to technical writing.
Worked on a full-stack application to accurately fit users to a CPAP mask based on an image or 3D mesh. This involved development of a client side application (React) that inferfaced with a Postgres database in AWS as well as local hardware like a Bellus3D stereo camera. Also responsible for developing/evaluating custom and off-the-shelf model architectures for better sizing performance.
Contributed to a cross-functional machine learning research effort related to segmentation and analysis of biopsy images. Responsible for building ML infrastructure such as data pipelines (train / validation / test), utilities (data cleaning), and visualizations. Tested off-the-shelf networks for efficacy.
Developed a full-stack interal application to provide user-friendly access to hundreds of gigabytes of robot logs. Responsible for improving the aesthetics of the user interface, adding functionality like a 3D map viewer, and distilling developer requirements into feature roadmaps.
Collected and analyzed logs from a car-mounted positioning system (GPS, IMU, etc). Resposible for data collection and producing analysis of key performance metrics like GPS coverage over time.
This low cost, remote controlled rover is meant to be used in a search and rescue context to identify potential victims and assist first responders. It was developed as a senior design project at the University of Pittsburgh alongside (Eric Sherman, Tej Lalvani, Joey Black, and Stanley Galloway)
A research project with Rutgers Univsersity to produce a machine learning model that can accurately predic the progression of Oral Epithelial Dysplasia to oral cancer.
A small robot that recognizes and greets you! This was an undergraduate project for a design class with a timeline of ~ 1 month.
This is toy in the style of a 'BopIt' but with a custom racing theme. Hand designed, printed, and assembled, it puts the EE in speed. Completed alongside Eric Sherman and Tej Lalvani for Junior Design at the University of Pittsburgh.
# no posts yet - thinking is hard.