Software Engineer, Content Safety
Core
Building global-scale defensive architectures and pipelines to protect users from offensive, sensitive, or harmful content in UGC and GenAI products.
Role type
Mid-level Software Engineer (ML Infrastructure & Safety)
Builds
Global-scale content safety pipelines and defensive architectures for UGC and GenAI products
Domain
Internet safety, Generative AI, Natural Language Processing, Large-scale distributed systems
Deliverable
production ML models
Required skills
Python, Java, C++, Machine Learning, Large Language Model (LLM) architecture, transformers, model training, model inference, model deployment, model evaluation, optimization, data processing, debugging, adversarial incentives analysis, abuse vector identification, distribution dynamics analysis, Service Level Objectives (SLO) management, technical debt management, bug reduction
Preferred skills
Safety-adjacent domain experience (factuality, product policy, responsible AI), managing safety for UGC or GenAI products, designing global-scale defensive architectures, managing technical debt and SLO breaches, deep understanding of transformer activations and efficient large-scale training
Technologies
Python, Java, C++, transformers, LLMs
Responsibilities
Design and deploy global-scale defensive architectures and pipelines, manage safety for UGC or GenAI products, contribute to foundational models, maintain high operational standards by mitigating SLO breaches
Seniority
Mid-level, hands-on IC
