Senior Machine Learning Engineering Manager, Safety AI Systems

Written by the T&S Network from Roblox's posting.

Leads ML engineering for multimodal content safety at Roblox, building models that detect abusive behavior across avatars, images, video, audio, text, code, and their composites. The role manages a team of ML engineers and works with product, policy, and Trust & Safety teams to detect and mitigate abuse before it harms the community.

• Harassment and bullying

What you'd do

• Own technical direction and execution of the Multimodal Safety AI system's detection of harmful content at scale.
• Lead and mentor a team of ML engineers, including growing their skills and handling conflict.
• Turn long-term safety goals into a staged roadmap with measurable value at each step.
• Guide architecture of large vision language and other large-scale models, making them production-ready.
• Align technical decisions across Safety and other teams, building consensus among stakeholders.
• Shape the multimodal safety machine learning plan alongside partners in Product, Data Science, Policy, Design, and Operations.
• Track emerging AI/ML and content moderation techniques and adapt safety approaches to new challenges.

What it asks for

• Five or more years running large machine learning systems in live production environments.
• Track record of launching ML models from scratch into production.
• Two or more years working directly with vision language or other foundation models.
• Experience with microservice architectures and distributed systems, such as cloud services and scalable data pipelines.
• Ability to set clear success metrics and strategic roadmaps.
• Hands-on enough to dive into code and architecture when needed.

Our best summary of the posting, and it can be wrong. The employer's posting is what counts: read it in full on Robloxnorth_east before you apply. If we got something wrong, report this listing.