Model Infrastructure Engineer Graduate (TikTok Recommendation Architecture) - 2026 Start (BS/MS)

Responsibilities

About the Team
The Recommendation Architecture team is responsible for building up and optimizing our recommendation system's architecture to provide the most stable and best experience for our users. As a New Graduate, you'll join a high-impact team focused on optimizing Large Language Models (LLMs) and large-scale recommender models optimization on GPU platforms. You'll build and scale AI infrastructure that powers state-of-the-art models in production.

We are looking for talented individuals to join our team in 2026. As a graduate, you will get opportunities to pursue bold ideas, tackle complex challenges, and unlock limitless growth. Launch your career where inspiration is infinite at TikTok.

Successful candidates must be able to commit to an onboarding date by end of year 2026. Please state your availability and graduation date clearly in your resume.

Candidates can apply to a maximum of two positions and will be considered for jobs in the order you apply. The application limit is applicable to TikTok and its affiliates' jobs globally. Applications will be reviewed on a rolling basis - we encourage you to apply early.

Responsibilities
- Optimize model performance and memory efficiency on GPU-based systems.
- Collaborate with research and infra teams to deploy high-throughput training and inference pipelines.
- Develop tools and libraries to accelerate deep learning workloads at scale.
- Analyze system performance (e.g., GPU profiling, kernel analysis, throughput tuning).

Qualifications

Minimum Qualifications:
- Final year or recent graduate with a a background in Computer Science, Electrical Engineering, or other related field.
- Solid programming skills in C++/CUDA/Trition/Python.
- Familiarity with GPU architecture and distributed training is highly desirable.

Preferred Qualifications:
- Experience building production-grade training and inference systems for large-scale models.
- Hands-on experience optimizing Large Language Models (LLMs), including memory efficiency, latency, and throughput improvements.
- Knowledge of distributed training frameworks (e.g., NCCL, Horovod, DeepSpeed, FSDP) is a plus.