North Carolina Institute for Climate Studies (NCICS)
|Data Scientist Intern
Highlights
Designing and implementing ETL data pipelines using AWS services (SQS, S3, Lambda) to automate and streamline data ingestion, transformation ensuring efficient access to climate datasets for public health and climate research applications.
Performing data cleaning and preprocessing using Pandas and NumPy, including handling missing values, standardizing data formats, and removing outliers to enhance data quality.
Implemented zero-shot text classification using a BERT-based DistilBERT model to categorize real-world Parent Crisis Hotline messages into thematic topics, enabling scalable analysis without labeled training data.