Vithursan Thangarasa is a machine-learning researcher focused on neural-network compression, large-scale foundation models, sparse training, and efficient inference. He leads inference ML work at Cerebras Systems and has previously worked at Cerebras, Uber AI Labs, Tesla, and other engineering organizations.
Vithursan Thangarasa is a machine-learning researcher focused on neural-network compression, large-scale foundation models, sparse training, and efficient inference. He leads inference ML work at Cerebras Systems and has previously worked at Cerebras, Uber AI Labs, Tesla, and other engineering organizations.
2025 — Present
Leads inference machine-learning research and engineering focused on efficient deployment of large language and multimodal models. Works on model compression, speculative decoding, sparsity, and hardware-software co-design.
2022 — 2024
Conducted research on sparse training, neural-network compression, and large-scale language models. Authored and contributed to Cerebras research and technical publications.
2020 — 2022
Worked on machine-learning systems, sparse model training, and efficient foundation-model development.
2019 — 2019
Contributed to machine-learning research and engineering projects involving efficient neural-network training.
