TikTok
Software Engineer Intern (Recommendation Infra, Performance Efficiency) - 2027 Summer
San Jose, California · Internship · Posted Oct 28, 2026
- Education
- Currently pursuing a Bachelor's degree in Computer Science or a related technical discipline; currently pursuing a Master's degree in the same fields is preferred.
About the job
Build and optimize high-performance infrastructure, serving systems, and automation pipelines for TikTok’s large-scale global recommendation systems.
What you'll do
- Design and optimize recommendation serving systems for efficiency, reliability, scalability, and serving quality.
- Find and resolve performance bottlenecks and improve infrastructure and compute cost efficiency.
- Build automation workflows and pipelines for recommendation-system data processing and development.
What they ask for
- Currently pursuing a Bachelor's degree in Computer Science or a related technical discipline.
- Strong programming skills in at least one systems programming language, such as C++, C, Go, or Java.
- Experience building scalable backend, distributed, infrastructure, or high-performance online systems.
- Solid understanding of data structures, algorithms, operating systems, networking, and distributed system fundamentals; experience with performance analysis, debugging, reliability improvement, or large-scale service optimization.
Nice to have
- Experience with recommendation, search, advertising, machine learning, or large-scale online serving infrastructure; or optimizing high-throughput, low-latency C++ services.
- Experience with profiling, performance tuning, large-scale data processing systems, or AI-agent/LLM-based automation workflows.
Tools
C++, C, Go, Java, Flink, Spark, Kafka, LLMs
AI focus
The team is developing AI-native and agentic workflows, including closed-loop systems that observe, diagnose, plan, execute, verify, and improve recommendation infrastructure. Related preferred experience includes automation using AI agents and LLMs.
Model-written summary · the original posting is authoritative