Senior ML Platform Engineer — Real-Time AI for Kids

Spin Master; New York (NY); Hybrid; USD 166,000 - 221,000; Full time

Benefits offered by this job

Flexible hours

Growth opportunities

Comprehensive benefits

Other perks

Job summary

Spin Master’s Hapiko unit seeks an ML Engineer to own the architecture and performance of the machine learning pipeline powering Stickerbox. You will scale technology for a global product launch and make foundational architectural decisions for a consumer hardware product. Hybrid work: 3 days weekly in Long Island City.

You will own latency, deploy and manage model inference, and lead migrations while ensuring kid-safe compliance and scalable serving. Strong production experience is essential.

Qualifications

• Experience deploying diffusion or similar generative ML models in production with real latency constraints.
• Experience profiling a generation path to optimize for performance and cost.
• End-to-end production ownership experience, including deployment, capacity planning, observability, and on-call support.
• Advanced knowledge of ML infrastructure, inference systems, and production software architecture.

Responsibilities

• Own end-to-end latency from input to printed sticker, and profiling to remove bottlenecks.
• Design and manage model deployment, versioning, and rollback strategies for hardware devices.
• Optimize batching, caching, and GPU utilization across image generation, speech recognition, and character models.

Skills

ML deployment in prod

Performance profiling

End-to-end ownership

ML infrastructure knowledge

Product-driven design

Edge inference

Diffusion models

Job description

Spin Master’s Hapiko unit seeks an ML Engineer to own the architecture and performance of the machine learning pipeline powering Stickerbox. You will scale technology for a global product launch and make foundational architectural decisions for a consumer hardware product. Hybrid work: 3 days weekly in Long Island City.

You will own latency, deploy and manage model inference, and lead migrations while ensuring kid-safe compliance and scalable serving. Strong production experience is essential.