AI Infrastructure Intern

Tencent; Singapore; On-site; SGD 39,000 - 61,000; Full time

Job summary

Tencent seeks students for an internship/early career role to design and implement high-performance orchestration systems for large-scale model training and real-time inference. You will contribute to scheduling across global GPU/CPU clusters and improve framework efficiency.

Ideal candidates are pursuing CS/IS degrees, have experience with ML orchestration frameworks, and possess strong backend software, Linux, and teamwork skills.

Qualifications

• Currently enrolled in an undergraduate or graduate degree program in Computer Science, Information Systems, or related fields.

Skills

Backend languages (Golang, Python, C/C

Distributed systems

Linux familiarity

ML frameworks (Ray, TFX, Kubeflow)

Teamwork / collaboration

Education

Undergraduate or graduate degree in CS/IS or related fields

Tools

Ray

TFX

Kubeflow

Job description

Who We Look For

• Design and implement high-performance orchestration systems to automate the deployment, scaling, and lifecycle management of large-scale model training and real-time inference services
• Develop scheduling algorithms to optimize the utilization of global GPU/CPU clusters. Focus on improving resource allocation efficiency and managing the complexities of heterogeneous hardware
• Participate in development of training-related requirements (incremental training and API for online learning) and enhance framework efficiency
• Participate in development of pipelines for offline-to-online synchronization. Ensure strict data consistency and minimize model-update latency to ensure the most current intelligence is served
• Collaborate with cross-functional teams to ensure platform features are available, secure, and compliant for a global user base.
• Currently enrolled in an undergraduate or graduate degree program in Computer Science, Information Systems, or related fields
• Experience in machine learning system practice and open-source ML orchestration frameworks (e.g. Ray/TFX/Kubeflow)
• Proficiency in backend software design, development, and deployment practices with at least one of the following programming languages: Golang, Python, C++, or Java
• In-depth knowledge of distributed system principles (e.g., consistency protocols like Paxos or Raft, distributed locking, and caching strategies)
• Familiarity with Linux operating system and common system tools
• Strong ownership, customer-oriented values, and integrity demonstrated
• Good programming discipline, fast-learning ability, and teamwork skills
• Prior internet industry work or internship experience is a plus