Aug 2024 – Jan 2025 · 5 mos · San Francisco, California, United States

• Built a bimanual non-humanoid robot with 2 other engineers to build a robotics startup in CPG manufacturing
• Achieved autonomous movement using TensorFlow and the Action Chunking with Transformers (ACT) reinforcement learning model using data collected via teleoperation on top of CAN servo drivers we wrote ourselves.
• Crawled and scraped web data to prospect for sales leads using CommonCrawl, Python, Anthropic's Claude Sonnet, Scrapy, Beautiful Soup, Selenium and rotated proxies with AWS lambdas.

Apr 2023 – Jun 2024 · 1 yr 2 mos

• Created Leanbakers, a protein powder brand for baking

Oct 2022 – Jul 2024 · 1 yr 9 mos · Los Angeles, California, United States

• Built a content creation recommendations engine on topics & products for brands and content creators using clickstream data from over 25 million users and 63 billion events with daily ingestion of over 250 million daily events using DocumentDB, AWS CDK, AWS lambda, AWS IAM, AWS SQS, and more.
• Built a product that supplied daily Youtube data to over 14 million brands and content creators which included data of their accounts, performance of their videos, their comments, and subscriptions, using Dagster, Databricks, AWS S3, Python, and Apache Kafka, to name a few.

Sep 2021 – Sep 2022 · 1 yr · Natick, Massachusetts, United States

• Product that analyzed sensor data at the world’s largest factory and logistics companies such as Amazon, Walmart, and Apple, via the edge intelligence team.
• Saved customers average of 3 hours in factory/logistics automation downtime by building backend API endpoints for a data analytics feature written in Python utilizing a Flask framework and running queries for a PostgreSQL database
• Introduced a novel method of prototyping new features that cuts down pilot development resources (time and developers) by 80% by building infrastructure connecting our data pipelines allowing the use of streamlit applications written in Python
• Assist product customers from the largest logistics companies of the world with major conveyor line blocks through on call support debugging the proprietary software stack on a Linux platform

Jan 2019 – Apr 2019 · 3 mos · Goddard, Maryland

• Optimized and converted static NASA Earth Science data manipulation tutorials to interactive Jupyter Notebooks implemented in Docker images and written in Java, Python, C, C++ depending on data needs that were viewed and used by over 10,000 users
• Condensed the time to create new content by 1 hour by automating and optimizing several parts of the creation process which allowed increased scalability
• Reduced storage requirements by over 40% by cutting out unnecessary dependencies

Oct 2018 – Dec 2018 · 2 mos · Merced, California Area

• Optimized processes required to create virtual reality content for the WAVE at UC Merced
• Automated and documented VR content creation workflow using Python and Jupyter for better readability by non- programmer docents and photographers/videographers
• Troubleshot and fixed hardware and software issues pertaining to the creation and presentation of large-scale Stereo Panoramas and dense point clouds

Jun 2018 – Aug 2018 · 2 mos · Los Angeles

• Built software that can predict and simulate radar scattering with a high degree of accuracy to optimize development workflow saving over 24 hours of processing time per image
• Built a radar imaging program from scratch using Python to image a simulated target being hit by a pulse modulated radar signal
• Developed Python code to integrate radar clutter backscatter coming from the ground
• Read and translated previously written Fortran code into C, for updating and integrating into proprietary repository written in C, which simulated scattering by a dielectric sphere in a radar simulation
• Co-authored paper on radar imaging for the IEEE Research and Applications of Photonics in Defense (RAPID) conference, Miramar Beach, FL, August 20, 2019.