Software Engineer, Content Safety

Overview

This role at Google involves developing systems that ensure online content safety by tackling offensive, sensitive, or harmful materials. The position supports Google’s core products, leveraging advanced machine learning techniques and collaborating with DeepMind to enhance responsible AI implementations. The software engineer will contribute across the tech stack, focusing on safety-critical content systems and infrastructure.

Key Responsibilities

• Create, implement, and scale systems for content safety, including classifiers, vector databases, and multimodal machine learning models supporting business-critical and generative AI products.
• Develop and sustain robust distributed systems and pipelines for content processing optimized for reliability and throughput on both server and device platforms.
• Manage model lifecycle activities including training, evaluation, deployment, and iterative improvement via automated feedback loops.
• Design advanced agent-based workflows and heuristics to deeply understand and preempt complex abuse behaviors and threat vectors.
• Lead agile development efforts to identify, respond to, and mitigate emerging abuse patterns to maintain user engagement and product safety in a dynamic threat environment.

Candidate Qualifications

Minimum requirements:

• Bachelor’s degree or equivalent relevant experience.
• At least 2 years of programming experience in Python, Java, or C++.
• Minimum 1 year of experience working in core machine learning domains such as generative AI, NLP, computer vision, speech/audio, reinforcement learning, recommendation systems, or ML infrastructure.
• At least 1 year experience handling ML infrastructure tasks like model training, inference, deployment, evaluation, optimization, data processing, and debugging.

Preferred qualifications:

• Experience related to safety topics including factuality, product policy, or frameworks for responsible AI.
• Exposure to safety management in user-generated content (UGC) or generative AI products with an understanding of adversarial tactics, abuse mechanisms, and viral dissemination dynamics.
• Background in architecting and scaling defensive systems and pipelines that meet strict service level objectives (SLOs) on a global scale.
• Strong conceptual grasp of machine learning and large language model architectures, particularly transformer models and aspects of efficient large-scale training and deployment.
• Proven track record in managing technical debt, reducing bugs, and preventing SLO violations to uphold operational excellence.

Additional Information

Applicants currently authorized to work in Singapore and not requiring visa sponsorship from Google will be prioritized. Google emphasizes equal opportunity employment and affirmative action, welcoming applications without bias to protected characteristics. Accommodations are available for applicants with disabilities upon request.