Dwarak Talupuru is a Senior Staff Research Engineer based in London with eight years of experience specializing in LLM pretraining, distributed training, and performance engineering. At Cohere he leads training performance engineering and hardware-aligned architecture design, driving cost reductions for frontier models and enabling large-scale MoE training across thousands of GPUs. His background spans building TPU+JAX+Ray distributed frameworks, sparse-network research, and score-based generative models from research roles at Carnegie Mellon and startups. Known for shipping production-scale pretraining systems (Command R / R+ and Command A workstreams), he blends deep algorithmic insight with pragmatic systems optimization. Colleagues value his ability to translate hardware constraints into architecture and scaling strategies that materially lower training cost and time. He brings a rare combination of hands-on distributed systems craft and academic rigor applied to real-world LLM training challenges.
7 years of coding experience
6 years of employment as a software developer
BTech - Bachelor of Technology, BTech - Bachelor of Technology at Manipal Institute of Technology
high school, Computer science, high school, Computer science at velammal vidhyashram
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.
Request Free Trial
Dwarak Talupuru - Senior Staff Research Engineer at Cohere