Staff Data & ML Infra Engineer — Scale Video Pipelines

Cantina Labs; Greater London; On-site; GBP 150,138 - 195,179; Full time

Benefits offered by this job

Medical insurance

Paid time off

Parental leave

401(k) retirement plan

Lifestyle spending account

Office lunch and snacks

Equity

Job summary

Cantina Labs is hiring a Member of Technical Staff to build and scale data pipelines for our large video generation models. You will own annotation workflows, dataset curation, and preprocessing tools that directly improve model quality.

You’ll collaborate with research and engineering to turn experiments into scalable systems, ensuring data is clean, ready for training, and delivered efficiently within our Kubernetes-based infrastructure.

Qualifications

• 3+ years of experience in machine learning, data pipelines or related engineering roles on large-scale multimodal or video systems.
• Strong programming skills in Python and building reliable data processing pipelines for ML workflows.
• Hands-on experience preparing training data including parsing, filtering, and dataset curation using cloud platforms.

Responsibilities

• Build and maintain data pipelines for large video generation models including ingestion, parsing, filtering, preprocessing, and dataset curation at scale.
• Design and run annotation workflows across MTurk, Prolific, and Mechanical Turk with quality control.
• Train and evaluate smaller models used for data filtering and preprocessing in the ML pipeline.
• Partner with research and engineering teams to scale experimental workflows into repeatable systems.
• Own data quality across the pipeline and continuously improve tooling and processes.
• Build internal tools to prepare datasets and monitor outputs for model development.
• Drive large pipeline projects from inception to completion, including dataset upgrades.
• Work within a Kubernetes-based training infra to ensure data is prepared and delivered to training clusters.
• Profile and optimize inference scripts used in preprocessing for time and cost efficiency.

Skills

Python

Data pipelines

ML preprocessing

Kubernetes

MTurk/Prolific workflows

Cloud computing

PyTorch

Cross-functional collaboration

Large-scale data handling

Education

CS/ML degree

Tools

AWS S3

DynamoDB

RunPod

Job description

Cantina Labs is hiring a Member of Technical Staff to build and scale data pipelines for our large video generation models. You will own annotation workflows, dataset curation, and preprocessing tools that directly improve model quality.

You’ll collaborate with research and engineering to turn experiments into scalable systems, ensuring data is clean, ready for training, and delivered efficiently within our Kubernetes-based infrastructure.