Data Scientist – Patient Predictive Modelling

Confidential Client; Hyderabad; Hybrid; INR 1,800,000 - 2,600,000; Full time

Job summary

Confidential Client in Hyderabad, India seeks a Data Scientist – Patient Predictive Modelling to design and deploy AI/ML models on Real-World Data, forecasting clinical outcomes and informing intervention windows.

The role requires a graduate degree (Master’s preferred) with strong ML/DL, Python and experience with US pharma data systems, in a hybrid work setup in Hyderabad.

Qualifications

• Graduate degree (Master's preferred) in Health Informatics, Biostatistics, Computer Science, or Data Science.
• Strong experience in Machine Learning / Deep Learning along with knowledge on US Pharma data systems.
• Advanced Python capability with Scikit-Learn, XGBoost.
• Experience adapting deep learning architectures (PyTorch/TensorFlow) for time-series data.
• Experience manipulating large datasets using SQL, PySpark, and Databricks.
• Explainable AI (XAI).

Responsibilities

• Build and deploy predictive models (Survival Analysis, Markov Chains, RNNs) to map patient journeys and disease progression.
• Act as SME on US healthcare data structures, extracting, cleaning, and transforming unstructured data.
• Design algorithms to predict patient risks, including adverse drug events and hospitalization likelihood.
• Apply unsupervised clustering and representation learning on multimodal datasets to discover clinical subpopulations.
• Ensure feature-attribution mapping (SHAP, LIME) for regulatory stakeholders.

Skills

Machine Learning

Deep Learning

Python

Time-series

Explainable AI

SQL

PySpark

Education

Master's degree

Tools

Scikit-Learn

XGBoost

PyTorch

TensorFlow

Databricks

SHAP

LIME

Job description

Job Title: Data Scientist – Patient Predictive Modelling

As a Patient Predictive Modeler, you will use cutting-edge AI/ML to extract clinical insights from vast, multi-modal Real-World Data (RWD) systems. You will architect predictive models focused on tracking longitudinal patient data to forecast clinical outcomes, identify optimal windows for therapeutic intervention, map patient stratification, and simulate disease progression trajectories. This role directly impacts commercial deployment strategies.

Key Responsibilities

• Build and deploy predictive models (Survival Analysis, Markov Chains, Recurrent Neural Networks) to map complex patient journeys, lines of therapy, and disease progression.
• Act as a Subject Matter Expert (SME) on massive US healthcare data structures, extracting, cleaning, and transforming unstructured data from diverse health systems.
• Design algorithms to predict patient risks, including adverse drug events, hospitalization likelihood, treatment non-adherence, and clinical trial dropout rates.
• Apply unsupervised clustering and representation learning on multimodal biological and phenotypic datasets to discover distinct clinical subpopulations for precision medicine.
• Ensure all clinical and patient-level predictions feature clear feature-attribution mapping (SHAP, LIME) so global medical and regulatory stakeholders can trust model outputs.

Requirements

• Graduate degree (Master's preferred) in Health Informatics, Biostatistics, Computer Science, or Data Science.
• Strong experience in Machine Learning / Deep Learning along with knowledge on US Pharma data systems.
• Advanced Python capability alongside machine learning packages (Scikit-Learn, XGBoost).
• Experience adapting deep learning architectures (PyTorch/TensorFlow) for time-series or sequential data processing.
• Experience manipulating large datasets utilizing SQL, PySpark, and Cloud computing workspaces like Databricks.
• Explainable AI (XAI).

Work Arrangements

• Full-time position based in Hyderabad, Telangana, India.
• Work schedule: Hybrid (2-3 days Work from Office).