AI Operations Engineer

About the role

We are seeking an AI Operations Engineer to build and operate the large-scale data systems that power modern AI training and evaluation pipelines. The role combines deep data engineering expertise with a strong understanding of AI workloads, focusing on ingestion, transformation, quality assurance, lineage, and high-throughput delivery of data to training jobs across diverse modalities.

Key Responsibilities

• Design and operate large-scale data pipelines supporting AI training, evaluation, and continual improvement workflows
• Build ingestion systems for diverse modalities including text, image, audio, video, and structured signals
• Implement data cleaning, deduplication, filtering, and quality assurance at petabyte scale
• Develop dataset versioning, lineage, and provenance tracking systems suitable for reproducible training
• Build high-throughput data loading systems that maximize GPU utilization during training
• Design storage architectures balancing cost, throughput, and latency across data tiers
• Implement labeling workflows, active learning pipelines, and human-in-the-loop data improvement systems
• Drive observability of data quality, drift, and pipeline health across the AI data estate
• Optimize cost and performance through compression, format selection, and caching strategies
• Document data systems, schemas, and operational procedures for broad internal use
• Stay current with AI data infrastructure research and emerging open-source tools

Required Qualifications

• Bachelor’s or Master’s degree in Computer Science or a related field
• Six or more years of data engineering experience, with significant work supporting ML or AI workloads
• Strong proficiency in Python and at least one JVM or systems language
• Deep experience with modern data processing frameworks such as Spark, Ray, or Beam
• Hands-on experience operating petabyte-scale storage and pipeline systems
• Strong understanding of distributed systems, data modeling, and storage formats
• Experience with dataset versioning, lineage, and reproducibility for ML workflows
• Familiarity with high-throughput data loading for accelerator-based training
• Excellent communication and cross-functional collaboration skills

Preferred Qualifications

• Experience with multimodal datasets at large scale
• Familiarity with data quality tooling and dataset evaluation methodology
• Exposure to privacy-preserving data systems and regulated data handling
• Open-source contributions to data infrastructure projects

How To Apply

For immediate consideration, please send your resume to Harry@bvteck.com or contact us at (908)676-4399. Learn more about Bright Vision Technologies at www.bvteck.com. Bright Vision Technologies is an Equal Opportunity Employer.