Staff Engineer - Foundation Models & Pretraining

Hark; San Jose (CA); On-site; USD 180,000 - 450,000; Full time

Job summary

Hark, based in San Jose, California, is seeking a talented individual to join the Omni team focusing on large-scale pretraining systems and foundation models.

This role emphasizes innovation in multimodal AI experiences beyond text, with responsibilities including optimizing training infrastructures, collaborating across teams, and improving model performance through new data practices and evaluation frameworks.

The position offers a competitive salary in the range of $180,000 - $450,000 annually based on experience.

Qualifications

• Proven track record of enhancing neural network performance.
• Strong experience in large-scale distributed training.
• Deep understanding of LLM and multimodal pretraining.

Responsibilities

• Drive research in large-scale LLM and multimodal pretraining.
• Develop optimized data pipelines for pretraining.
• Design efficient training strategies for foundation models.

Skills

Large-scale neural network performance

Distributed training (Megatron, DeepSpeed)

Data-driven experimentation

High-performance computing systems

Job description

Hark, based in San Jose, California, is seeking a talented individual to join the Omni team focusing on large-scale pretraining systems and foundation models.

This role emphasizes innovation in multimodal AI experiences beyond text, with responsibilities including optimizing training infrastructures, collaborating across teams, and improving model performance through new data practices and evaluation frameworks.

The position offers a competitive salary in the range of $180,000 - $450,000 annually based on experience.