Obico, formerly The Spaghetti Detective, has a mission to build smarter 3D printers through AI and computer vision. I came on to TSD as an early stage co-founder and helped grow the business into one of the most popular pieces of paid software in 3D printing. At the time of this post TSD's AI powered failure detection algorithm had been run on over 45 million hours of 3D printing and was the top paid app on OctoPrint.

Transferred to the Mobile Vision Team after roughly 2 years full-time at Google. The Mobile Vision Team is a research group within Google AI (formerly Research & Machine Intelligence). The focus of this team is improving computer vision capabilities in key verticals with a particular emphasis on running modern neural networks in compute and energy constrained environments. Contributions by this team, such as MobileNets, have helped enable the rapidly growing edge vision space.During my time on the team, we grew from around 15 team members to over 100. At the time of my departure I was a co-lead for the full Mobile Vision Team and the Technical Lead for four sub-teams: Vision Mining Toolkit (large scale web data mining), Logo Recognition, Artwork Recognition, and Food Nutritional Understanding.Co-founded, advised, or contributed to the following publicly announced projects:-Cloud Vision API-Google Lens-Nutrition5k-Google [X] Computational Agriculture-Visual Positioning Service-Google Vision Kit

Did a temporary rotation working full-time embedded within Google [X]. During this time I was helping to advise the Computational Agriculture robotics team on how to improve their computer vision and edge artificial neural network systems.

Started my full-time work at Google on the Brand Formats Team. On this team I was building responsive front end ad units for high end brand advertisers.Co-created a new system for automatically generating rich responsive advertisement formats from an arbitrary set of advertiser assets. This novel approach generated an $XX M ARR for Google in the first year of deployment. We submitted and were issued a patent for this technique.Was the Technical Lead for a computer vision project during 20% time. During this time we created a novel indoor navigation system for visually impaired users. This project won 1st place in an internal "Innovation Week" competition, spurred the creation of a dedicated team within accessibility to continue the work, was highlighted at the internal company wide TGIF meeting, and helped motivate my transfer to the Mobile Vision research team. This was eventually released, after years of work by other talented people, as Project Guideline.

Worked remote on campus at UCLA promoting the Google brand and products. Helped UCLA with implementing the transition plan to have GMail as their primary mail server for students, faculty, and staff. Placed in the top five of hundreds of student ambassadors around the world in the yearly competition to accomplish the most impact on your campus.

Was privileged to serve as an officer and eventually president of the Computer Science Honor Society at UCLA. Our organization was responsible for hosting around 40 different events per school year helping students with mentorship, academic tutoring, industry networking, technical interview prep, and more.During my year as President we piloted a program where officers and inductees were required to take turns volunteering as tutors, providing unlimited, free, drop-in tutoring in most lower division courses. As of 2020 this program was still a tradition of UCLA UPE.

I worked for the summer of 2012 as a Software Engineer Intern. I continued my work in the fall as a part time Junior Software Engineer after returning to school. Work done is confidential; however, in the subsequent acquisition by Raytheon they stated publicly the following description of the work that the company did: "tactical intelligence, surveillance and reconnaissance, secure tactical communications and cybersecurity across a broad spectrum of globally dispersed platforms and communications networks."

I was a Software Engineering Intern doing front end work using JavaScript, CSS, Ajax.

I assisted with research in general purpose GPU programming. The goal was to find domain specific tasks where transferring execution to the GPU using CUDA could provide run time benefits that exceed the transfer time.I started with part-time work during the school year, focused on accelerating anti-virus scanning. This work was supervisor by Professor Glenn Reinman and was a collaboration with Symantec.Following my part-time work, I was selected for a full time role for the summer of 2010. For this work I was supervised by Professor Luminița Vese. My work during the summer focused on accelerating a 3-dimensional MRI image processing pipeline. My work resulted in approximately a 100x increase in run time speed relative to the baseline. This was made possible by slightly modifying the proximity filters to cut off a small contribution from the long tail (e.g. in Gaussian or Poisson filters) to enable a sliding window transfer approach. Required work optimizing the balance between memory transfer and computation on the GPU using low level CUDA control.

Assisted in building kits and lesson plans for teaching basic robotics and computer vision to undergraduate computer science students. Work was supervised by Professor Rahul Simha and sponsored by a National Science Foundation Grant.