Aaron T

Specializes in ML, AI, LLM-driven systems with background building scalable & data-driven apps across healthcare, biomedical, NLP domains. Highly proficient in Python & R with hands-on experience developing CNNs, RNNs, LSTMs, Transformers, XGBoost, random forests, and models for prediction, classification, and time-series analysis. Strong background in data preprocessing, feature engineering, explainable AI, survival analysis, dimensionality reduction, and transforming structured/unstructured data into actionable insights and scalable solutions.

Skilled in cloud-native development and deployment using AWS, incl. building automated pipelines and microservices.

Data Scientist Intern at Hypahub (2023-01 – 2023-03)

• Designed evaluation methodologies for BioGPT-generated medical guidance, enabling systematic benchmarking against GPT-2 and PubMedBERT and supporting model selection decisions for healthcare NLP workflows.
• Conducted pre-training on 15M abstracts from NCBI DB; demonstrated a 38.5%-39.8% performance improvement over GPT-2 and PubMedBERT on related tasks. Tackled problems in design reviews with team lead.

Webmaster and Designer at LGCC (2021-02 – 2022-12)

• Redesigned website architecture to improve content accessibility and user navigation efficiency. Created a management dashboard with data visualizations to track site traffic, ensuring data-driven improvements.
• Communicated and worked closely with management to ensure periodic updates met their requirements fully.

User Interface Designer at Reddit (2020-01 – 2021-01)

• Applied user behavior analytics and A/B testing insights to improve engagement and enhance accessibility outcomes.
• Integrated A/B testing results into design processes, reducing support tickets by 25%. Interacted and worked closely with clients to test site design elements.

Master of Science in Data Science – Northwestern University (2025-12)

Bachelor of Science in Informatics & Data Science – University of Massachusetts, Amherst