ML Data Engineer

About the Company

Recraft is an AI tool for professional designers, illustrators, and marketers, setting a new standard for excellence in image generation. Over 3 million users across 190+ countries have produced hundreds of millions of images using Recraft.

Responsibilities

• Develop and maintain data-ingestion pipelines to source and prepare large-scale image and text datasets
• Own the end-to-end flow from raw data to ready-to-train artifacts, including filtering and validation
• Operate and improve the Kubernetes-based data-pipeline framework
• Manage S3-style object storage layouts, lifecycle, and throughput
• Build tooling for pipeline observability, including progress visualization and metrics
• Collaborate with ML engineers to align datasets with training needs

Requirements

• Strong Python fundamentals and production-ready code
• Solid hands-on Kubernetes experience with containers and distributed processing
• Proven track record with unstructured data, specifically images at scale
• Experience developing data-ingestion or parsing tools for public sources
• Proficiency with S3/object storage
• English proficiency at B2+ level

Preferred Qualifications

• Familiarity with ML workflows and PyTorch
• Experience with image quality scoring or captioning pipelines
• Experience with DAG/workflow visualizations
• DevOps fluency with Docker and CI/CD