Summary
Gabriel Dumitrescu is a Staff Research Scientist with eight years of experience at the intersection of generative AI and production-scale systems, currently focusing on video transformer models and talking avatars. He helped build Hollywood-grade generative video tooling at Metaphysic.ai—pioneering diffusion-based face swapping, neural rendering, motion transfer and bias mitigation—and later designed distributed training pipelines with model sharding and sequence parallelism at BRAHMA AI. A startup founder who turned computer vision research into a funded mobile product, he blends hands-on ML research with product development, cloud deployment and performance optimization. He mentors and hires technical teams, having led recruitment that added ten senior engineers and refined interview processes. Comfortable across NLP, 3D neural rendering and real-time video pipelines, he pairs technical depth with an eye for UX and ethical model design.
8 years of coding experience
6 years of employment as a software developer
High School Diploma, Computer Science, High School Diploma, Computer Science at Tudor Vianu National High School of Computer Science
Master of Science - MS, Artificial Intelligence, Faculty of Mathematics and Computer Science, Master of Science - MS, Artificial Intelligence, Faculty of Mathematics and Computer Science at University of Bucharest
Start-up Accelerator, Start-up Accelerator at Innovation Labs 2022
Romanian, English, French