Data Engineer: Real-Time Lakehouse for Large-Scale Recs

TIKTOK PTE. LTD.; Singapore; Sur place; SGD 120 000 - 180 000; Plein temps

Avantages offerts par ce poste

Stock options

Yoga and fitness

Paid leave

Flat organization

Career growth opportunity

Résumé du poste

TikTok is seeking talented graduates to join our Arch-Data Ecosystem team, building offline and real-time data storage for large-scale recommendation, search, and ads at a billion-user scale. You will design data architectures, collaborate on storage, and optimize data pipelines for efficiency and cost.

Successful candidates will contribute to a unified training data base and state system, with opportunities to tackle bold ideas and advance in a rapid-growth tech environment.

Qualifications

• Candidates completing or recently completed a PhD in AI, CS, CE or software-related field.
• Experience building large-scale distributed systems, preferably storage, streaming, or ML infra.
• Familiarity with Flink internals, state management, connectors or UDFs.
• Knowledge of Lakehouse tech: Paimon, Iceberg, Delta Lake, or Hudi, incl. schema evolution and snapshots.

Responsabilités

• Build unified infra for training data base and training/inference state for multimodal models in search, recommendation, and ads.
• Collaborate on data lakes, caching, distributed computing, and GPU IO to reduce costs and improve iteration efficiency.
• Own components of distributed storage and processing stack from file formats to metadata management.
• We seek graduates to pursue bold ideas and grow within the team.

Formation

PhD in Artificial Intelligence

PhD in Computer Science

PhD in Computer Engineering

PhD in Software Development

Outils

Apache Flink

Apache Paimon

Iceberg

Delta Lake

Hudi

PyTorch

Description du poste

TikTok is seeking talented graduates to join our Arch-Data Ecosystem team, building offline and real-time data storage for large-scale recommendation, search, and ads at a billion-user scale. You will design data architectures, collaborate on storage, and optimize data pipelines for efficiency and cost.

Successful candidates will contribute to a unified training data base and state system, with opportunities to tackle bold ideas and advance in a rapid-growth tech environment.