Neerav Chaudhary

AI/ML Engineer and Data Scientist with hands-on industry and research experience developing LLM-powered tools, predictive models, and computer vision systems for healthcare. Improved detection accuracy and reduced false positives with custom churn-prediction solutions, and published deep learning research on brain atrophy quantification. Aims to use applied AI skills to support impactful, data-driven outcomes in healthcare environments.

Artificial Intelligence Research Intern - Deep Strat

(2023-08 - 2023-11)

• Built churn and loan-default prediction models (Random Forest, CNN), designing a custom loss function that jointly optimized false-positive rate and profit to maximize detection reliability, reducing false positives by 25% while maintaining stable AUC
• Prototyped an LLM-powered ticket assistant that performed sentiment and intent analysis on support emails and chat logs, auto-classified and routed issues to departments via vectorized data retrieval, and generated SOP-grounded response drafts to standardize resolution quality
• Authored a 50-page technical reference spanning the full ML/AI stack - from regression fundamentals to LLM architectures, failure modes, and bias propagation - used to diagnose specific causes of model compliance and ethical failures for stakeholders
• Contributed technical analysis and stakeholder discussions that informed a jointly authored conference paper on AI safety and responsible deployment practices, presented at Nullcon

AI Researcher, Medical Imaging - Manipal University Jaipur

(2022-08 - 2022-12)

• Designed and evaluated CNN architectures for dementia severity classification on the ADNI longitudinal MRI dataset, testing depth-width trade-offs and combining models in an ensemble framework that improved accuracy from 78% to 85%
• Implemented an OpenCV-based method to quantify neural atrophy from MRI scans, automating extent-of-damage estimation to reduce the time required for manual assessment of disease progression (published in IEEE ICRITO 2022)
• Built image-processing workflows to segment affected brain regions and extract structural features, linking imaging outputs to clinically meaningful indicators of atrophy severity across patient cohorts aimed at helping clinicians prioritize patients for further evaluation

M.S. - Industrial / Engineering Management - San Jose State University (2024-08 - 2026-05)

Bachelor's - Computer Engineering - Manipal University Jaipur (2018-08 - 2022-06)