I am an AI research engineer with over a decade of experience building and maintaining large-scale codebases for foundational models. I currently work on mulitmodal efforts at Microsoft AI, including MAI-Image-1 and internal video models. Before that, I contributed to Imagen 3 (text-to-image), Veo 1 and 2 (text-to-video) at DeepMind. Earlier in my career, I led neural network design efforts at Google Brain, serving as the Tech Lead of Flax. I also did quite a bit of theoretical research in knowledge representation & reasoning, and compositionality in language.
Throughout my career, I’ve worked across diverse areas of AI, collaborating with researchers and developers from various backgrounds and cultures. I thrive in dynamic environments, wearing multiple hats, and I’m passionate about fostering a positive and inclusive work culture where everyone feels empowered.
You can find more on my resumé (pdf), Github, or Google Scholar.
Feel free to reach out at marcvanzee [at] gmail [dot] com.