Time Series Machine Learning Intern

Summary

PHM Society, based in Andover, MA, is seeking a Time Series Machine Learning Intern for Summer 2026. The intern will develop data science products using time series machine learning algorithms and work closely with offer managers and usecase squads.

Responsibilities

• Develop Python-based time series machine learning algorithms for fleetwide IoT data, including anomaly detection and classification
• Create frameworks for prototyping and benchmarking machine learning algorithms and tools on data across fleets of similar industrial assets
• Select appropriate datasets and data representation methods
• Run machine learning tests and experiments to support the product development
• Develop and test strategies for training and retraining machine learning systems
• Research and survey industrial reports and technical papers on latest industrial trends, case studies, models and methodologies

Skills

• The candidate must be currently enrolled in a masters or doctoral graduate program in a related discipline such as Mathematics, Statistics, Physics, Engineering, Data Science or CS
• Demonstrable experience in full stack data science, including developing and deploying machine learning models in a production environment
• Experience working with time series data is required
• Experience exploring and training models with big datasets, using cloud technologies (especially Databricks)
• Proficient in scientific Python (NumPy, SciPy, scikit-learn, Pandas, XGBoost), SQL, and version control (GitHub)
• Ability to formulate hypotheses, draw conclusions and deliver results
• Excellent verbal and written communication and interpersonal skills
• Ability to present findings to stakeholders and communicate results to all colleagues
• Prior experience with Python-based libraries for time series, such as SKTime, tsfresh, Ruptures
• Familiarity with time series machine learning algorithms/ frameworks like ROCKET, Matrix profile techniques, contrastive learning, and architectures like InceptionTime and other 1D Convolutional Neural Networks
• Familiarity with time series foundational models, like MOMENT, TimeGPT and Granite

Qualifications

Must Haves

• The candidate must be currently enrolled in a masters or doctoral graduate program in a related discipline such as Mathematics, Statistics, Physics, Engineering, Data Science or CS
• Demonstrable experience in full stack data science, including developing and deploying machine learning models in a production environment
• Experience working with time series data is required
• Experience exploring and training models with big datasets, using cloud technologies (especially Databricks)
• Proficient in scientific Python (NumPy, SciPy, scikit-learn, Pandas, XGBoost), SQL, and version control (GitHub)
• Ability to formulate hypotheses, draw conclusions and deliver results
• Excellent verbal and written communication and interpersonal skills
• Ability to present findings to stakeholders and communicate results to all colleagues

Nice to Haves

• Prior experience with Python-based libraries for time series, such as SKTime, tsfresh, Ruptures
• Familiarity with time series machine learning algorithms/ frameworks like ROCKET, Matrix profile techniques, contrastive learning, and architectures like InceptionTime and other 1D Convolutional Neural Networks
• Familiarity with time series foundational models, like MOMENT, TimeGPT and Granite

Benefits

• Overtime pay
• Recognition programs