REMOTE Sr. ML/AI Engineer

Our customer is seeking a highly experienced Machine Learning / AI Engineer (8+ years) to design, develop, and operationalize predictive models that identify track nonconformities across a metropolitan subway system.

This role will work within a Google Cloud Vertex AI environment, leveraging multi-modal sensor data (vibration, audio, location) captured via Pixel-based hardware kits on revenue cars, along with ground-truth defect data from MTA's Hexagon (HxGN) system. The ideal candidate will play a critical role in advancing the TrackInspect application by delivering production-grade ML solutions for predictive maintenance and anomaly detection.

Key Responsibilities

Perform advanced feature engineering on sensor and operational datasets to identify patterns associated with known track defects in HxGN

Refactor and enhance existing prototype ML models for pilot and production-scale deployment

Utilize Vertex AI Workbench to train, retrain, version, and track ML models, experiments, and performance metrics

Conduct hyperparameter tuning, model validation, and optimization to improve nonconformity prediction accuracy

Deploy models for daily batch inference, integrating with pipeline-driven sensor data ingestion systems

Implement model monitoring, alerting, and feedback loops using Track Inspector inputs within HxGN

Design and maintain CI/CD pipelines and multi-environment deployment strategies for ML lifecycle management

Build and maintain BigQuery-based prediction views, including location metadata, prediction outputs, and user feedback fields

Evaluate model outputs against known defect datasets and present findings in model evaluation sessions with MTA stakeholders

Collaborate cross-functionally with engineering, data, and business teams to improve model performance and usability

8+ years of professional experience in Machine Learning, AI, or Data Science roles

Strong proficiency in Python (e.g., scikit-learn, TensorFlow, PyTorch) and SQL

Hands-on experience with Google Cloud Vertex AI (Workbench, training pipelines, model deployment)

Deep understanding of

• Feature engineering techniques
• Model training and validation frameworks
• Hyperparameter optimization methods

Experience designing and deploying production-grade ML systems

Experience building data pipelines and CI/CD workflows for ML models