ML Trust & Safety Engineer for Kids' AI Sticker Printer

Spin Master; New York (NY); On-site; USD 137,000 - 182,000; Full time

Benefits offered by this job

Growth opportunities

Flexible hours

Innovation & Fun

Comprehensive benefits

Other perks

Job summary

Spin Master is seeking a dedicated Trust Safety Engineer to own the moderation pipeline for Hapiko, a kid-focused sticker printer powered by generative AI. You will ensure safety from spoken prompts to printed images and collaborate with ML and backend teams.

You will build evals, tune classifiers, run red teaming, maintain incident runbooks, and help meet data retention and deletion obligations while balancing protection and user creativity.

Qualifications

• Applied safety engineering experience on a product with real users.
• Built evals or classifiers for harmful content, not just vendor APIs.
• Strong Python software engineering for production moderation code.
• Ability to balance precision/recall and set thresholds with cost considerations.
• Good judgment under pressure.

Responsibilities

• Own end-to-end content moderation from user input to printed output.
• Maintain and grow safety eval suites and determine readiness of models/thresholds.
• Improve classifiers and decide where in-house vs third-party models fit.
• Tune moderation to reduce over-blocking while maintaining protection.
• Run ongoing red teaming against models and pipelines.
• Maintain review tooling for flagged content and parent reports.
• Own the incident response runbook for safety issues.
• Maintain data retention and deletion compliance."
• Collaborate with ML and Backend/Platform engineers to bake safety into new features.

Skills

Safety engineering

Evals/classifiers

Python

Precision recalls

Decision making

Job description

Spin Master is seeking a dedicated Trust Safety Engineer to own the moderation pipeline for Hapiko, a kid-focused sticker printer powered by generative AI. You will ensure safety from spoken prompts to printed images and collaborate with ML and backend teams.

You will build evals, tune classifiers, run red teaming, maintain incident runbooks, and help meet data retention and deletion obligations while balancing protection and user creativity.