Postdoctoral Fellow

Application and inquiries should be submitted by e-mail to Dr. Kovacheva, vkovacheva@bwh.harvard.edu, with "Application for AI/ML postdoc position" in the subject line. Along with your CV, please include a cover letter describing previous research, research interests, and future goals. Please provide contact details for 3 references.

Position Description

A postdoctoral fellowship is immediately available in the research group led by Dr. Vesela Kovacheva in the Department of Anesthesiology at Harvard Medical School / Brigham and Women's Hospital.

We are seeking a highly motivated, collaborative individual passionate about developing predictive models that enhance patient safety and prevent adverse pregnancy outcomes. You will utilize multidimensional clinical datasets-including waveform signals (e.g., ECG, EEG), genetic data, and imaging-to create predictive algorithms targeting critical maternal outcomes such as hypertensive crises, hemodynamic instability, hemorrhage, and ICU admission. You will also contribute to developing NLP-based and time-series models and integrating these models directly into clinical practice.

Our state-of-the-art data platform provides access to billions of clinical data points from over 300,000 patients, enabling groundbreaking research with immediate translational potential.

You will be part of a diverse, multidisciplinary team of data scientists, clinicians, and researchers in a stimulating academic environment, with ample opportunities for collaboration across all Mass General Brigham hospitals, Harvard Medical School, the Program in Medical and Population Genetics at the Broad Institute, and industry partners.

Qualifications

Qualified candidates should have:

- Ph.D. in a quantitative discipline such as data science, bioinformatics, computer science, biomedical engineering, or a related field (preferably completed within the last 3 years).

- Strong programming skills in Python, with proficiency in deep learning frameworks (PyTorch or TensorFlow) and other data-processing libraries (Polars)

- Demonstrated expertise in machine learning, deep learning, explainable AI (XAI), and analysis of time-series data.

- Experience in medical imaging analysis with AI,

- Familiarity with ultrasound image preprocessing, feature extraction, and multimodal data fusion using relevant libraries (MONAI, ITK, OpenCV, scikit-image).

- Experience implementing AI interpretability methods (SHAP, attention visualization, etc).

- Excellent oral and written communication skills, meticulous attention to detail, and strong professional integrity.