Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

I am an AI/ML engineer with about four years of experience delivering production-grade machine learning and generative AI systems across healthcare, hospital billing, and insurance domains. I enjoy turning messy data into actionable insights and scalable AI solutions. I’ve built patient risk stratification models, medication adherence predictors, and end-to-end analytics pipelines on AWS, plus NLP and multimodal models. My work emphasizes measurable clinical and business impact and close collaboration with care teams to drive better outcomes and efficiency.…I am an AI/ML engineer with about four years of experience delivering production-grade machine learning and generative AI systems across healthcare, hospital billing, and insurance domains. I enjoy turning messy data into actionable insights and scalable AI solutions. I’ve built patient risk stratification models, medication adherence predictors, and end-to-end analytics pipelines on AWS, plus NLP and multimodal models. My work emphasizes measurable clinical and business impact and close collaboration with care teams to drive better outcomes and efficiency.

Nitish Kumar Manthri

AI Developer, Data Scientist, AI Engineer, +2





I am an AI/ML engineer with about four years of experience delivering production-grade machine learning and generative AI systems across healthcare, hospital billing, and insurance domains. I enjoy turning messy data into actionable insights and scalable AI solutions. I’ve built patient risk stratification models, medication adherence predictors, and end-to-end analytics pipelines on AWS, plus NLP and multimodal models. My work emphasizes measurable clinical and business impact and close collaboration with care teams to drive better outcomes and efficiency.…I am an AI/ML engineer with about four years of experience delivering production-grade machine learning and generative AI systems across healthcare, hospital billing, and insurance domains. I enjoy turning messy data into actionable insights and scalable AI solutions. I’ve built patient risk stratification models, medication adherence predictors, and end-to-end analytics pipelines on AWS, plus NLP and multimodal models. My work emphasizes measurable clinical and business impact and close collaboration with care teams to drive better outcomes and efficiency.

Available to hire

I am an AI/ML engineer with about four years of experience delivering production-grade machine learning and generative AI systems across healthcare, hospital billing, and insurance domains. I enjoy turning messy data into actionable insights and scalable AI solutions.

I’ve built patient risk stratification models, medication adherence predictors, and end-to-end analytics pipelines on AWS, plus NLP and multimodal models. My work emphasizes measurable clinical and business impact and close collaboration with care teams to drive better outcomes and efficiency.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Intermediate

Julia

Intermediate

Work Experience

Data Scientist at CVS Health

June 1, 2024 - Present

Built patient risk stratification models on claims, pharmacy, and clinical data using Python, CatBoost, Pandas, and SQL, improving high-risk case handling for 15,000+ members annually with 85% accuracy. Developed a medication adherence prediction system using ensemble learning, temporal feature engineering, and semantic search with FAISS, enabling care teams to act on 2,100 non-adherent patients monthly and increasing adherence interventions by 25%. Designed end-to-end AWS-based analytics pipelines (S3, Glue, Athena, Redshift) that reduced reporting turnaround from 7–8 hours to under 2 hours and cut annual costs by $250K. Implemented NLP pipelines with clinical BERT and SpaCy to extract insights from physician notes, discharge summaries, and care plans, accelerating utilization review and case resolution workflows by 40%. Established multimodal disease progression models combining LSTM for temporal data and CNNs for imaging/document data, supporting early-stage risk identification in

Data Scientist at KPIT

June 1, 2020 - August 1, 2022

Engineered patient payment risk models on hospital billing and claims data, identifying 6,800 delayed accounts monthly to enable timely finance interventions. Built predictive payment adherence system using Gradient Boosting and CatBoost with semantic embeddings for document search, reducing follow-up call time by ~1,200 hours per quarter. Built ETL and feature engineering pipelines with PySpark, Dask, and Airflow processing 1.5 TB monthly, cutting preprocessing time from 3 days to under 12 hours. Applied NLP via Spark NLP to extract insights from billing notes and insurance documents, automating case review and improving resolution time. Designed anomaly detection with Isolation Forest and Gradient Boosting to flag unusual billing trends, generating alerts preventing ~$220k revenue losses per month. Created patient segmentation with K-Means and PCA to optimize outreach; built interpretable AI pipelines with SHAP and LIME; productionized pipelines with TensorFlow, Databricks, and Kuber