Looking for a job

V User

@vuser14

Data Scientist at DenuoSource India building healthcare cost-of-care models that identified 22% of high-cost outlier claims.

India

Message

At DenuoSource India, I develop risk-adjusted cost-of-care models for Amplify Health, identifying 22% of high-cost outlier claims and improving payer negotiations by about 11%. I also built 30-day readmission risk models to support early intervention.

I engineered more than 20 clinical and demographic features that lifted model performance by 12–15%, and streamlined provider scorecards across 2,000+ hospitals, reducing manual reporting effort by 80%.

At 10bet, I designed a churn prediction model that increased retention by 70%, developed deposit forecasting that increased revenue opportunities by 60%, and created a fraud detection model that reduced false negatives by 80%.

Earlier, I worked on cross-channel fraud detection at Paymaya (Maya) and industrial machine learning at Ntwist. At Ntwist, my deep learning and optimization work helped cut operational costs by 60%, and I built a computer vision module for bitumen flotation level detection.

Experience

Work history, roles, and key accomplishments

Current

Data Scientist

Current

DenuoSource India Pvt Ltd

Aug 2024 - Present (2 years 2 months)

Developed risk-adjusted cost-of-care models and readmission risk models using PySpark, improving payer negotiations and supporting value-based care. Engineered clinical features and streamlined provider scorecards across 2,000+ hospitals.

PySpark Machine Learning Healthcare Analytics Feature Engineering Data Modeling

Machine Learning Engineer

10bet

Jun 2023 - May 2024 (11 months)

Designed churn prediction and deposit forecasting models, increasing retention by 70% and revenue opportunities by 60%. Developed fraud detection model reducing false negatives by 80%.

EvalML XGBoost Random Forest Churn Prediction Fraud Detection Forecasting

Data Science Consultant

Paymaya

Sep 2022 - Apr 2023 (7 months)

Built supervised ML models for cross-channel fraud detection using Feedzai, achieving 80% accuracy. Orchestrated Airflow DAGs to automate data pipeline and monitoring.

Feedzai Airflow AWS Sagemaker Redshift Fraud Detection MLOps

Python Developer – Machine Learning

Ntwist

Jul 2019 - Sep 2022 (3 years 2 months)

Developed deep learning models for industrial process optimization, achieving 95% accuracy and cutting operational costs by 60%. Built computer vision module for autonomous detection.

TensorFlow Keras Pyomo PCA Computer Vision Deep Learning Python