Inference Software Engineer: High-Performance Transformers

Etched.ai, Inc.; San Jose (CA); On-site; USD 120,000 - 180,000; Full time

Benefits offered by this job

Medical, dental, and vision packages

Housing subsidy of $2k per month

Relocation support

Various wellness benefits

Daily lunch + dinner in the office

Job summary

Etched.ai, Inc. is looking for candidates to support porting AI models to its innovative architecture in San Jose, California. Responsibilities include enhancing runtime environments and optimizing performance layers. Ideal applicants are proficient in C++ or Rust and have experience with distributed software systems. Benefits include a housing subsidy, relocation support, and comprehensive health packages. Join a fully in-person team that values engineering and research collaboration.

Qualifications

• Experience with complex distributed software systems and performance-sensitive applications.
• Ported applications to non-standard accelerator hardware or platforms.
• Developed low-latency applications using user-space networking.

Responsibilities

• Support porting state-of-the-art models to our architecture.
• Build and enhance Sohu’s runtime for multi-node inference.
• Analyze performance profiling and optimize communication layers.

Skills

C++ proficiency

Rust proficiency

Understanding of Linux internals

Familiarity with PyTorch

Familiarity with JAX

Job description

Etched.ai, Inc. is looking for candidates to support porting AI models to its innovative architecture in San Jose, California. Responsibilities include enhancing runtime environments and optimizing performance layers. Ideal applicants are proficient in C++ or Rust and have experience with distributed software systems. Benefits include a housing subsidy, relocation support, and comprehensive health packages. Join a fully in-person team that values engineering and research collaboration.