Senior AI Engineer (Trust and Safety)
Core
Own end-to-end AI models for content safety (image, chat, voice) to detect prohibited content and automate moderation at massive scale.
Role type
Senior AI Engineer (Trust and Safety)
Builds
Production moderation models (classifiers, LLMs, VLMs) deployed on self-hosted GPU fleets for user prompts and generated content.
Domain
Generative AI, Trust and Safety, Content Moderation
Deliverable
production ML models
Required skills
End-to-end model ownership, fine-tuning and evaluation of LLMs/VLMs, adversarial robustness, active learning, data labelling strategies, cost/latency optimization, Python, GPU infrastructure management
Preferred skills
Image moderation (Vision classifiers/VLMs), Text moderation (distilled LLMs), Synthetic data generation, Adversarial content detection, High-scale deployment experience
Technologies
Python, PyTorch, TensorFlow, Hugging Face, Kubernetes, Docker, AWS, Self-hosted GPU clusters
Responsibilities
Define detection requirements and build classifiers/LLMs for prohibited content, Train and evaluate models on GPU fleets, Deploy models into production and monitor performance, Automate retraining loops and content review workflows, Optimize models for accuracy, latency, and cost, Collaborate with labellers to improve data quality
Seniority
Senior, hands-on IC