Candidate
Employer
PhD Intern – Database Systems Kernel
Are you applying to the internship?
Job Description
Research Scientist Intern (Compute Platform – Database Acceleration) – 2026 Start (PhD) | TikTok
The Tone:
This is a PhD internship at TikTok, a global company known for its short-form mobile video platform. The role is within the Data Platform team, specifically the DB Boost team, which is dedicated to building next-generation database systems. This position offers the opportunity to contribute to the core infrastructure that underpins data-driven decision-making across TikTok’s business products. This internship provides hands-on learning, community-building, and collaboration with industry experts.
The TL;DR
• Role: Internship
• Location: In-person
• Pay: $36–$45 hourly
• Team: Data-Data Platform-Compute Platform – DB Boost team, reporting to Data Platform team
• Mission: Building the next generation DBMS kernel aiming for best-in-class query latency and hardware utilization.
• Tech Stack: Velox, DataFusion, DuckDB, Polars, Clickhouse, Spark, Presto
What You’ll Actually Do
• Design: Design and develop kernel modules or core features for cloud-native distributed analytical database systems.
• Research: Conduct research for cutting-edge technology evolvement in database and data management and identify potential future directions.
• Problem-Solve: Solve unique and complex problems on large-scale, high-performance cloud infrastructure.
• Collaborate: Partner with different teams on building backend support for cloud native database service.
The Must-Haves
• Background: Doctorate Degree (currently enrolled or obtained) in Computer Science, a related STEM, or quantitative field. Career level: Student.
• Experience: Experience with at least one system programming language.
• Skills: C++, Rust, database kernel, query processing, query optimization, storage systems, hardware acceleration, cloud native infrastructure.
• Bonus: Research or deep working experience in OLAP databases and data management systems, experience with open-source and industry databases (Velox, DataFusion, DuckDB, Polars, Clickhouse, Spark, Presto), contributing to open-source communities, creativity, and quick problem-solving capabilities.
