Design and implement custom GPU/accelerator kernels to maximize performance for next-generation AI workloads. Collaborate with researchers to translate algorithmic advances into efficient, production-ready code.
Mindbeam
3 Remote Job Openings at Mindbeam
Develop pipelines for post-training tasks including fine-tuning, evaluation, and model compression. Implement scalable systems for model deployment and optimization while collaborating with researchers to validate results in production.
The primary mission involves designing and optimizing large-scale pre-training systems to power generative AI models. This includes building scalable pre-training pipelines, implementing distributed training strategies across hardware, and developing monitoring systems for reliability.
