ML Data Engineer — Scalable Pipelines & Curation

Cantina; Singapore; On-site; SGD 90,000 - 150,000; Full time

Benefits offered by this job

Competitive salary and equity

Personal time off and holidays

Health insurance

Global travel insurance

Monthly stipend: $500 (~S$635)

Home office equipment

Job summary

Cantina Labs is seeking an ML Engineer in Singapore to build and scale data pipelines for large-scale video and multimodal data used in model training. You will own the end-to-end pipeline from raw content to curated, training-ready datasets with emphasis on speed, reproducibility, and cost-efficiency.

You will work with curation and modeling teams to evolve dataset recipes, implement VLM-based captioning and filtering, and optimize data storage, processing, and deduplication at scale.

Qualifications

• Experience designing and scaling large-scale data systems for ML.
• Proficient in distributed data processing (PySpark or Ray).
• Experience with containerization and orchestration (Docker, Kubernetes).
• Familiar with cloud storage/compute across AWS, GCS, or Azure.

Responsibilities

• Design and scale distributed data pipelines for preprocessing, dataset generation, and refreshed datasets.
• Own workflow orchestration, job scheduling, monitoring, and failure recovery for large-scale data processing jobs.
• Implement and maintain containerized pipeline infrastructure using Kubernetes or equivalent systems.
• Optimize cloud storage and movement across providers for cost, throughput, and efficiency.
• Define and implement best practices for dataset storage, versioning, caching, retention, and access patterns.
• Design curation pipelines to select, filter, and retain content for model training.

Skills

Data pipelines

PySpark

Ray

Airflow

Docker

Kubernetes

Python

Video/Multimodal ML

CLIP/embedding filtering

FFmpeg/OpenCV

Education

BS/MS in CS/Math/ML

Tools

Airflow

Docker

Kubernetes

PyTorch

TensorFlow

Job description

Cantina Labs is seeking an ML Engineer in Singapore to build and scale data pipelines for large-scale video and multimodal data used in model training. You will own the end-to-end pipeline from raw content to curated, training-ready datasets with emphasis on speed, reproducibility, and cost-efficiency.

You will work with curation and modeling teams to evolve dataset recipes, implement VLM-based captioning and filtering, and optimize data storage, processing, and deduplication at scale.