Machine Learning Engineer - TikTok BRIC Live Integrity - Singapore

TikTok Live is a world-leading live streaming platform serving hundreds of millions of users globally. By joining us, you will step into one of the industry's most challenging environments, characterized by massive scale, high concurrency, and complex global adversarial scenarios.

This role offers immense potential for professional growth and the opportunity to apply cutting-edge AI technologies in real-world applications. You will protect the safety and financial security of our global community, setting new benchmarks for industry risk control.

• Risk Identification & Governance: Lead the detection and mitigation of ecosystem/content risks and financial risks on TikTok Live.
• Model Development: Design and optimize high-precision, low-latency risk detection models utilizing massive multimodal data (text, image, audio, and behavioral sequences).
• Adversarial Defense: Analyze black-market tactics and attack patterns to design robust defense strategies and respond rapidly to emerging threats.
• Advanced AI Application: Drive the adoption of cutting-edge technologies, including AI Agents, LLM Fine-tuning, and Multimodal Models, to enhance automated reasoning and decision-making capabilities in complex risk scenarios.

• Bachelor’s degree or higher in Computer Science, Mathematics, Statistics, or a related field.
• Algorithmic Foundation: Solid background in machine learning and deep learning; proficiency in classification, clustering, and sequence modeling algorithms (e.g., XGBoost, DeepFM, Transformers).
• Engineering Skills: Strong coding proficiency in Python, C++, or Go. Experience with mainstream frameworks like TensorFlow or PyTorch. Experience with large-scale distributed systems is a strong plus.
• Strong logical thinking and data sensitivity. Demonstrated passion for risk control and adversarial challenges. Excellent communication and teamwork skills.

• AI Agent Experience: Proven track record in building AI Agents for complex task planning, tool use, or autonomous decision-making.
• LLM Expertise: Hands-on experience with LLM fine-tuning techniques such as SFT and RLHF; experience applying Large Language Models to vertical domains.
• Multimodal Capabilities: Deep understanding of multimodal representation learning, video understanding, or cross-modal retrieval.
• Domain Experience: Prior experience in community content safety, trust & safety (T&S), or anti-fraud.