IT Professional

Job Summary

Alzheimer’s disease is a leading cause of death in the United States, and scalable digital tools for prevention and early detection are urgently needed. Our team is building machine‑learning‑driven digital biomarkers from wearable sleep electroencephalography (EEG) to assess neurodegenerative changes before symptoms appear.

Examples of projects include:

• Brain‑aging phenotypes derived from EEG‑based digital features
• Cognitive‑decline trajectories linking sleep‑related physiology to performance
• Multimodal correlates, integrating EEG with neuroimaging and Alzheimer’s molecular/structural measures

We’re seeking a Part‑Time Data Scientist who can analyze real‑world datasets and build well‑documented, reproducible applied mathematical analysis pipelines in Python. You will drive exploratory data analysis, feature engineering, and model development, while applying solid documentation and coding practices that make results reliable and repeatable.

Candidates with backgrounds in Applied Mathematics who bring strong data‑analysis skills and practical ML experience are encouraged to apply. Equivalent real‑world experience is welcome.

• Perform exploratory data analysis on sleep EEG and related datasets; define targets, features, and baselines aligned with digital biomarker goals.
• Develop, evaluate, and compare models for classification/regression using sound validation design, including cross‑validation, regularization, calibration, and structured error analysis.
• Translate scientific questions into measurable model objectives and evaluation criteria, balancing performance with robustness and interpretability.
• Document experiments and results clearly; draft brief model cards and dataset summaries, including assumptions, limitations, and intended use.
• Build and maintain robust workflows to ingest, validate, and transform sleep EEG and related signals into analysis‑ready formats.
• Implement preprocessing and feature‑generation steps appropriate for wearable biosignals, with attention to data quality, artifact handling, and consistent labeling/alignment.
• Implement data quality checks and lightweight reporting to catch issues early (e.g., missingness, signal‑quality variation, label inconsistencies).
• Other duties assigned.

Remote – this role is eligible to work remotely, but the employee must be in the United States.

The University of Colorado Anschutz Medical Campus is a public education, clinical and research facility serving 4,500 students and a world‑class medical destination at the forefront of life‑changing science, medicine, and healthcare. CU Anschutz offers more than 42 highly rated degree programs through six schools and colleges, supported by $704 million in research awards in fiscal year 2023, creating an overall economic impact to the state of Colorado of $11.5 billion.

We are the single largest health‑professions education provider in Colorado, awarding nearly 1,450 degrees annually.

• Bachelor’s in Computer Science, Electrical/Computer Engineering, Data Science, Statistics, Applied Math, Neuroscience, or a related field from an accredited institution.
• One (1) year of professional IT project‑management experience (or substitution: an advanced degree may be substituted for experience on a year‑for‑year basis).

• Master’s degree in a related field.
• Applied math methods relevant to biosignals: coursework or project experience in time‑series analysis, statistical learning, signal processing, numerical optimization, inverse problems, or stochastic processes—especially applied to noisy real‑world data.
• Biosignal/EEG experience: exposure to EEG (sleep or otherwise) or related physiological signals (ECG, actigraphy, respiration); familiarity with common preprocessing/feature concepts (filtering, spectral features, artifact handling). Experience with tools such as MNE‑Python, YASA, or comparable MATLAB/R toolboxes is a plus.
• Practical model‑evaluation skills: experience designing train/test splits and cross‑validation for subject‑level/time‑series data; structured error analysis; calibration/thresholding…