Machine Learning Engineer, Predictive Maintenance
AssetWatch, Inc.1 month ago
Remote, United StatesSenior
Base Salary
$151k - $175k/yr
Responsibilities
- Extract and analyze vibration, temperature, electrical, image, and machine-context data to develop diagnostic features and fault signatures.
- Select, train, validate, compare, and deploy classical ML, deep learning, anomaly detection, time-series, signal-processing, hybrid, and rules-based models for machinery fault diagnosis and prognosis.
- Design end-to-end ML solution architectures covering data contracts, batch or near-real-time processing, feature computation, storage, orchestration, inference, monitoring, scalability, and cost.
- Build distributed data and feature pipelines over large sensor datasets using Spark, AWS services, Athena, and Parquet-based layouts.
- Design data mining, annotation, labeling, benchmark, validation, and regression workflows with condition-monitoring and domain experts.
- Develop modular, testable, maintainable Python software, shared libraries, model code, tests, documentation, and production workflows.
- Integrate LLM- and agent-based decision-support capabilities and use AI-assisted coding for ML and software development.
- Collaborate with MLOps and Data Engineering teams on deployment, CI/CD for ML, model versioning, infrastructure as code, containerization, monitoring, alerting, and production troubleshooting.
- Present model performance, experiments, tradeoffs, and recommendations to technical teams and stakeholders.
Requirements
- Master's or Ph.D. in Mechanical Engineering, Electrical Engineering, Computer Science/Engineering, or a related field.
- Proven predictive maintenance and condition-monitoring experience with signal-processing methods including FFT, Short Time Fourier Transform, Time Synchronous Averaging, Wavelet Transform, Spectral Kurtosis, Spectral Correlation, Envelope Analysis, and Hilbert Transform.
- Strong knowledge of classical and deep learning, supervised and unsupervised learning, anomaly detection, feature engineering, Artificial Neural Networks, CNNs, RNNs, LSTMs, attention mechanisms, and time-series modeling.
- Hands-on experience architecting and delivering production ML solutions on AWS, including services such as SageMaker, S3, Athena, AWS Glue, EMR, Lambda, Step Functions, ECS/Fargate, Timestream, Aurora/RDS, Amazon Bedrock, IoT Core, CloudWatch, and IAM.
- Experience with distributed processing using Spark or PySpark and efficient handling of large time-series and binary signal datasets.
- Experience designing expert-annotation workflows, label taxonomies, labeling tools, label quality control, and model-ready ground truth datasets.
- Strong Python software engineering skills with production code, Git, code review, unit/integration/regression testing, debugging, CI/CD, and reproducible development practices.
- Proficiency with NumPy, Pandas, SciPy, scikit-learn, TensorFlow, PyTorch, or similar scientific and ML/DL tools.
- Experience with Large Language Models, AI-assisted coding, and agentic AI tools or workflows.
- Working knowledge of SQL, including MySQL/Aurora and Athena.
- Experience with benchmark datasets, reviewed ground truth, false-positive and false-negative analysis, and repeatable evaluation or regression workflows is highly desirable.
- Physics-informed ML and experience with electrical/current signature analysis, ultrasound, or oil analysis are additional advantages.
- Strong communication and collaboration skills with domain experts, MLOps, Data Engineering, Product, and other stakeholders.
- Published research in high-quality journals or significant related applied work is a plus.
Benefits
- Remote-first work across locations in the United States and Ontario, Canada, with collaboration required during core working hours.
- Flexible work schedule.
- Stock options and comprehensive benefits including a retirement plan match.
- Unlimited PTO.
- Opportunity to work on impactful predictive-maintenance technology with a dynamic, growing team.
Tech Stack
Categories
About AssetWatch, Inc.
AssetWatch provides an end-to-end condition monitoring and predictive maintenance platform for industrial manufacturers, bundling wireless vibration/temperature sensors, connectivity, analytics software, and certified analysts. The service is sold by subscription with hardware and expert monitoring included, enabling quick, low-IT installation and trial programs for monitoring dozens of assets; founded in 2014 and headquartered in Dublin, Ohio, the company is privately held.