Research Engineer

📋 Description

• Orchestrate and optimize distributed training runs (DDP/FSDP)
• Build high-throughput data loading over TB-scale multimodal data
• Wrangle complex multi-stream time series and dataset versioning
• Build feature pipelines that are versioned and reproducible
• Write clean, well-tested research libraries for datasets and models
• Build evaluation harnesses that surface regressions automatically

🎯 Requirements

• 3+ years building ML systems or research infrastructure
• Expert-level Python and deep knowledge of PyTorch internals
• Track record of building or maintaining research libraries
• Experience with large unstructured and multimodal data pipelines
• Experience with experiment tracking and dataset versioning tooling
• Research fluency to read papers and reimplement components

🎁 Benefits

• Comprehensive health, dental, and vision insurance fully covered
• Relocation assistance and workspace setup stipend
• Monthly team dinners and twice-yearly off-sites
• Fully stocked kitchen with snacks
• Sponsorship for relevant conferences and professional development
• 12 weeks fully paid parental leave