Hybrid ML Infrastructure Engineer

at BlueSpace.ai

• Job title: ML Infrastructure Engineer
• Job type: Full Time
• Experience level: Junior, Mid level
• Salary: $150,000 - $200,000 per year
• Degree requirement: Bachelor's Degree
• Tech skills: Airflow, Cloud, Distributed Systems, Kubernetes, Node.js, Python, Spark
• Location requirements: Hybrid, Oakland, United States

About the role

• ML Infrastructure Engineer developing Cloud Data Infrastructure to support Assured AI for Autonomy. Designing and developing infrastructure to enhance Bluespace's APNT capabilities.

Responsibilities

• Take ownership of the infrastructure in support of developing and deploying Machine Learning models for Autonomous Vehicles
• Architect and deploy cloud and on-prem ML training and evaluation infrastructure
• Own the the data management pipelines, from ingestion and storage, to model training and evaluation that span vehicle compute, cloud, and on-prem
• Change model training code to take advantage of the better data storage techniques and formats you propose
• Evaluate and implement methods, software, and hardware for model deployment onto the test and production vehicles
• Develop systems and processes to improve transition of models from research to production while balancing cost
• Participate in model design, research and set requirements to model design that ensure their successful deployment
• Own and deliver projects end-to-end
• Optional: be able to hire, manage, or at least mentor other engineers who join this project when growth is needed

Requirements

• Experience in architecting and implementing data engineering solutions for a small engineering team / product (1-20 ppl)
• 2+ years of software engineering experience in any of the following: ML Infrastructure, Data Engineering, Platform Engineering, Distributed Systems
• Either existing experience with ML Infrastructure as described below, or strong expertise in non-ML Data infrastructure combined with a strong desire to learn ML Infra specifics
• Production ML experience with at least one of the following - (1). Model conversion and optimization for production (ONNX, TensorRT), (2) Model deployment on specialized hardware (e.g. Jetson), or (3) Model monitoring and MLOps
• Ability to programmatically access cloud services using Python, NodeJS, or equivalent
• Knowledge of or experience with data management solutions, such as - (1) Workflow orchestration pipelines (e.g. Argo, Airflow, Kubernetes) or (2) Managed large-scale data processing systems (e.g. Spark, Dataproc, Databricks)
• End-to-end ML pipelines (e.g. SageMaker, Vertex)