In this role at TikTok USDS JV, you will help build and operate the machine learning infrastructure behind hyper-scale feed and e-commerce recommendation systems. The position centers on reliable MLOps and data pipelines for real-time streaming, feature stores, model evaluation, and large-scale deep learning deployment with monitoring.
Working onsite in San Jose, you will collaborate with algorithm researchers and platform teams to move recommendation prototypes into zero-downtime production, while improving candidate generation and ranking for CTR and CVR serving.
What you’ll do
- Design, build, and maintain robust MLOps pipelines (including CI/CD for ML) for hyper-scale recommendation models supporting 10M–100M+ users, with continuous training, automated evaluation, and seamless deployment.
- Architect high-throughput real-time streaming data pipelines and feature stores, including offline and online feature synchronization to model dynamic user interest using short-term and long-term signals.
- Optimize production predictive models for candidate generation and ranking, focused on CTR and CVR.
- Improve inference performance through acceleration, model quantization, parallel computation, and latency reduction using C++/CUDA/TensorRT or comparable frameworks.
- Build automated tools for monitoring, debugging, and visualization to track feature drift, model performance degradation, data leakage, and GPU/CPU cluster utilization.
- Partner with algorithm researchers and platform teams to streamline the lifecycle of large-scale deep learning models, from experimental prototypes to zero-downtime production deployment.
Required qualifications
- Bachelor’s degree or higher in Computer Science or a related field.
- Strong programming and problem-solving ability.
- Basic understanding of MLOps principles, including data preprocessing, containerization (Docker), model tracking/versioning, or workflow orchestration.
- Motivation to address complex real-world distributed systems and machine learning operational challenges.
Technologies you’ll work with
- MLOps, CI/CD
- Docker
- C++, CUDA, TensorRT
- CTR, CVR
Compensation
- USD 122,574 - 259,200 per year
Benefits
- Day one access to medical, dental, and vision insurance
- 401(k) savings plan with company match
- Paid parental leave
- Short-term and long-term disability coverage
- Life insurance
- Wellbeing benefits
- 10 paid holidays per year
- 10 paid sick days per year
- 17 days of Paid Personal Time (prorated upon hire with increasing accruals by tenure)
- Eligible for additional discretionary bonuses/incentives and restricted stock units
Team and company context
You will join a group of applied machine learning engineers and data scientists focused on general feed recommendations and E-commerce recommendations. The team develops innovative algorithms and techniques to improve user engagement and satisfaction, turning creative ideas into solutions with business impact. They apply large-scale machine learning to solve real-world problems.
TikTok USDS Joint Venture LLC is dedicated to the safety and security of millions of Americans who create, discover, and connect with what they love on the apps operated by the company. The Joint Venture is established in compliance with an Executive Order signed by President Trump on September 25, 2025, and operates a comprehensive data privacy and cybersecurity program under defined safeguards to protect national security and secure U.S. user data, apps, and the algorithm. The company is shifting from a hybrid work model to a fully in-person schedule up to 5 days a week.
Preferred qualifications
- Experience in recommendation systems, online advertising, information retrieval, natural language processing, machine learning, large-scale data mining, or related fields.
Reasonable accommodation
USDS is committed to providing reasonable accommodations in recruitment processes for candidates with disabilities, pregnancy, sincerely held religious beliefs, or other reasons protected by applicable laws. To request assistance, contact: https://tinyurl.com/USDS-RA.
