Amlan Bhattacharya

Designed SQL schemas and complex queries to analyze transactional data and detect fraud patterns in highly imbalanced datasets. Integrated SQL with Python pipelines for feature engineering and model training, evaluating classifiers using precision and recall. Independently built and validated the full pipeline from data ingestion to model evaluation.

Maintained structured SQL tables for image metadata, preprocessing stages, and model experiment tracking. Used SQL to compare model performance across noise-reduction and segmentation techniques, analyzed class distribution and preprocessing impact, and logged model parameters and results in SQL-backed storage. Reviewed end results and cross-referenced findings against prior dataset analysis.

Created SQL tables for MRI scan metadata, augmentation strategies, and model outputs. Performed comparative analysis using GROUP BY and filtering to identify optimal configurations, and generated SQL-based reports to document experimental results. Reviewed results against prior dataset analysis to ensure correctness.

Built an NLP pipeline to classify text into positive, negative, and neutral sentiment categories. Stored raw text, processed tokens, and sentiment scores in relational SQL tables, and used SQL queries to analyze sentiment distributions across categories and time periods. Independently designed, built, and validated the full pipeline.