Projects

Making legal texts more accessible with ML models that detect plain vs. complex sentences.

Fine-tuned BERT models to capture emotions and sentiment in political communication.

Closing data gaps with backtranslation, paraphrasing, and embeddings for stronger NLP models.

Revealing hidden structure in political texts with clustering and embedding-based mapping.

Interpretable NLP systems with SHAP, attention maps, and robustness tests for fair and reliable predictions.

Designing and deploying NLP systems that work across languages, domains, and scripts with scalable pipelines.

Writing accessible articles on AI and technology trends, bridging research insights with public understanding.