AI Infrastructure Engineer
Spellbrushover 2 years ago
Responsibilities
- Design, implement, and operate next-generation inference architecture for models powering mobile, web, and other applications
- Develop infrastructure for running image-generation models across all platforms
- Collaborate with AI researchers and the AI team on state-of-the-art image-generation models
Requirements
- Experience with large distributed systems and on-premises and multi-cloud clusters
- Familiarity with Kubernetes, Kafka, NATS, and Redis
- Deep understanding of the tradeoffs and failure modes of introduced systems
- Excellent understanding of GPUs handling large workloads
- Experience deploying or optimizing GPU workloads end to end is a plus
- Comfort working on small, fast-paced teams
- Passion for anime, gaming, and generative AI
- Applicants should share previous work experience or a resume, GitHub, or portfolio
Benefits
- Equity compensation
- Employer-sponsored health, dental, and vision insurance
- Additional perks
- Preferred on-site collaboration in Tokyo or San Francisco
- Visa sponsorships available
Tech Stack
Categories
About Spellbrush
Spellbrush builds anime-focused video games and trains generative image models for artists and game production. The company is the team behind niji・journey, the anime image generator available through Midjourney. Founded in 2018 and headquartered in San Francisco with a research office in Tokyo, it operates its own GPU cluster to train models and develops in-house games with an emphasis on turn-based tactics.