About

Competition Robotics

I participated in the VEX Robotics Competition (now branded as the VEX V5 Robotics Competition) as a member of the world-renowned VEXMEN of Brandywine Robotics throughout my secondary schooling, first as a part of Team 80P VEXMEN: Pyro for the 2012–2013 season (Sack Attack), and after the team was split in two, as a part of Team 80Y VEXMEN: Cypher for the 2013–2014 (Toss Up) and 2014–2015 (Skyrise) seasons. VEXMEN: Cypher transitioned into high school with a number change to 81Y for the 2015–2016 season (Nothing But Net), and would continue participating that way for the 2016-2017 (Starstruck), 2017–2018 (In The Zone), and 2018–2019 (Turning Point) seasons, after which the team aged-out of competition.

Teams 80Y and 81Y VEXMEN: Cypher were highly successful. 80Y attended the 2015 VEX Robotics World Championship, and through a combination of luck, skill, and smooth-talking, managed to find itself a Middle School Tournament Finalist (and nearly winning the prestigious Excellence Award). After a gap of a year, 81Y qualified and returned to the VEX Robotics World Championship in 2017 by winning the High School Excellence Award at the Pennsylvania State Championship, of which it would return with an Engineering Award. 81Y attended the VEX Robotics World Championship for the last time in 2018 by qualifying as a Tournament Finalist of the Pennsylvania State Championship, and represented the team's most successful run: As the only team in the tournament to win every autonomous period it participated in, through sheer consistency it defeated the top-ranked seed in Research division to become a Division Champion and High School Tournament Finalist.

My primary role with VEXMEN: Cypher throughout its lifetime was, perhaps fittingly, being the team's dedicated programmer. The experience accumulated established a strong background in embedded programming, system design & integration, and controls.

Lockheed Martin Learning for Life Aviation

Hosted by volunteers of both former and then-current employees of the Sikorsky Coatesville plant, this program involved a variety of educational activities related to the field of aviation. The program's primary focus, however, was restoring a RotorWay Scorpion Too acquired by the American Helicopter Museum of West Chester, Pennsylvania, to displayable condition. After (many, many) hours of labor, the (now blue) Scorpion Too restored by the program is now on display, paired alongside a Model 133.

EleutherAI

I stumbled into joining EleutherAI as a contributor in February 2021, and would later become a core member by late March 2021. During my time volunteering, I would become maintainer and webmaster of the EleutherAI website and blog, redevelop its logo, and (most significantly) assist in the creation of GPT-NeoX-20B. After two years of dedicated contributorship and direct involvement in three published papers, various personal and interpersonal circumstances made my volunteering unsustainable and forced me to depart against my desire in March 2023.

Intel Innovators Program

As a long-time Intel graphics user perpetually jaded by NVIDIA's antics and intrigued by the emergence of a new competitor in the discrete graphics and accelerator market, I applied for and was granted access to the Intel Innovators program for my work on GPT-NeoX-20B in June 2023. This has led to my ongoing amateur involvement in Intel developer relations (notably as a moderator of the Intel DevHub Discord server), for which in return I've received hardware (such as an Arc A770 Limited Edition and a Lunar Lake Development Kit) to experiment with. I currently intend to continue in this role in some capacity until it becomes unsustainable or I am contractually obligated to cease.

Panverse Robotics

What came to be known as Panverse Robotics was a venture formed by me and a few other EleutherAI regulars in late-mid 2023 as an attempt to secure a Batch 2 Commercial AI Grant seed investment through a plan for advanced usage of neural fields for robotics. Despite making it far with our initial pitch, we were ultimately not selected and pivoted to a bootstrapping business plan built around the then-emerging 3D Gaussian Splatting (3DGS) concept, an initial phase of which was a consumer-focused 3DGS service marketed as Portality. Ultimately, the small team was unable to reach market fast enough before larger players with more resources entered, and with Portality still not fully launched before Q2 2024, the company reached the end of its run and was dissolved by the end of that calendar year.