I am Vice President of High-Performance AI at NVIDIA, a professor at Georgia Tech, and an engineer-researcher working across high-performance AI, multimodal AI, agentic AI and computer vision. My mission is to build the next generation high-performance, multimodal, and creative agentic AI systems that empower intelligence, creativity, and humanity.

Previously, I served as Chief Scientist at Picsart (2021-2025), where I built a global AI team from the ground up across research, engineering and product, and delivered AI tools used by 150M+ monthly users.  I also worked as Research Staff Member at IBM T. J. Watson Research Center, and professor at Oregon & UIUC.

Selected Highlights

• Leading efforts to build the next generation of high-performance, multimodal, and agentic AI systems that unify models, kernels, compilers, and hardware to accelerate intelligence and performance across the NVIDIA stack.

• Built and scaled a global AI organization at Picsart across research, engineering, and product, delivering large-scale multimodal AI systems (e.g., text-to-image models generating over 1 B images and tools used by 150 M+ monthly users) and a highly optimized GPU training and inference infrastructure.

• Open-source impact: NATTEN (Neighborhood Attention), StreamingT2V, Text2Video-Zero, Versatile Diffusion, OneFormer... —widely adopted across academia and industry; frequently GitHub trending; models downloaded millions of times. Hugging Face: #1 most-liked US university lab (Nov 2023).

• Research & competitions:  PI/Co-PI on multi-million-dollar projects across industry, academia, and government; papers across CVPR, NeurIPS, ICLR, ICCV, ECCV, TPAMI (including highly cited works); a dozen international AI competition wins (ImageNet, NVIDIA AI City Challenge, NIST TRECVID, DAC-System Design Contest, etc.).

• Honors: National Science Foundation CAREER Award; National Academy of Engineering Frontiers of Engineering (selected among 100 outstanding early-career engineers in the US).

• Industry recognition: Apple App Store “Trend of the Year: Generative AI” (2023); TIPA Best Consumer AI App (2023); IBM Research Accomplishment Award (2018).

• Leadership & mentoring: program chair, CVPR 2027; teach Computer Vision (UG/Grad); advised students now at NVIDIA, Google, OpenAI, Meta, Apple, Amazon, Tesla, etc.

For future collaborators: building high-performance multimodal & agentic AI—from kernels and algorithms to deployed systems and products. Email/DM welcome.

Recent Teaching & Services

• Computer Vision @ Georgia Tech (Fall 2023, Spring 2024, Fall 2024, Spring 2025)

• Workshop Chair: CVPR 2026

• Program Chair: CVPR 2027

• Area Chairs: CVPR 2021, 2022, ICCV 2021, 2025, ECCV 2020, 2022, ACM MM 2020, 2021, WACV 2024

• ArXiv Moderator for cs.CV: Computer Vision and Pattern Recognition , since 2024 

• Faculty Recruiting Chair: School of Interactive Computing, Georgia Tech

• AI Strategy Task Force, College of Computing, Georgia Tech