annabel ng

Hello! I’m Annabel, a 4th year EECS major at UC Berkeley excited about applying machine learning to solve challenging vision and language problems.
Currently, I’m spending the fall at Google as a software engineering intern on the YouTube Hardware Acceleration Team, where I’m exploring JAX-based quantized neural networks for Argos, YouTube’s custom video transcoding ASIC.
Highlights of my previous internships include:
- integrating a multi-agent conversational framework into one of 9 official AI Foundry customer templates on Microsoft’s AI Language team
- building ML tools to streamline and secure internal systems at Autodesk
- supporting the tap-to-pay rollout for the Boston MBTA with Cubic.
At Berkeley, I’m involved with computer vision research for autonomous F1 cars at ROAR as well as efficient image encoding research at the Redwood Center. I’m also passionate about improving the accessibility of machine learning education at Machine Learning at Berkeley, where I’ve hosted introductory ML workshops and bootcamps for underrepresented high school students.
I also enjoy curating playlists on Spotify, training for half marathons, or watching Studio Ghibli movies. If any of this interests you, feel free to shoot me an email!
news
Aug 9, 2025 | Contributed to Microsoft’s Conversational AI Accelerator, now with 50+ forks and 60+ stars ![]() |
---|---|
May 10, 2025 | Built N3CTAR - a pipeline for neural 3D cellular tessellated automata rendering ![]() |
Apr 23, 2024 | Published a paper in the 2024 Neuro-Inspired Computational Elements Conference ![]() |