Alina Lin Email & Phone Number

Kirkland, WA, US

Provo, UT And Seattle, WA, US

• AI and ML Innovations
• Developed a hybrid AI-human workflow that iteratively refines a classification schema, extracting actionable insights from free-text data, ensuring security, and maximizing explainability and user trust.
• Designed and implemented a comprehensive workflow integrating multiple LLMs and ML-based predictive models, combining unstructured and structured data to deliver an end-to-end solution that identifies actionable areas.
• Successfully completed two client POCs, gaining CEO-level attention.Research and Collaboration
• Led research analysis in collaboration with a partner company and three hospitals to integrate patient experience and clinical data, delivering impactful results within a one-quarter turnaround and culminating in a.
• Conducted strategic research projects to support internal company needs.Optimization and Algorithm Development

Providence, RI, US

• Improved personalized treatment prediction accuracy by over 30% byintegrating time series analysis with unsupervised machine learning to identify differential treatment trajectories with patient intake profiles
• Enhanced treatment success rates by 16% by conducting EHR-based clinical phenotyping to identify patients requiring longer lengths of partial hospital stays
• Optimized a patient utilization model to improve over 10% accuracy in treatment satisfaction prediction, translated into new guidelines for clinicians
• Improved psychodiagnostic assessment report writing speed by 300% by developing a Flask and GPT-based assessment report generator

• Built a scalable Shiny app based on time series analysis to predict patients’ treatment trajectories, improving prediction accuracy by 21% and serving as the prototype product of real-time clinical forecasting
• Established a cross-state behavioral science cloud network for data sharing and analysis targeting various stakeholders’ needs in mental healthcare

• Accelerated department research cycles by 5 times through providing customized data science consultations to 5 different labs
• Improved the positive predictability of minority mental health care access during the COVID-19 pandemic by 26% through addressing class imbalance issues with SMOTE sampling

• Developed a new probability graphical method combining bootstrapping to solve the historical difficulty in identifying moderators of relationships in high-dimensional cross-sectional data, explaining over 70% of the.

Stony Brook, NY, US

• Received >4.5/5 teaching evaluations for 2 advanced statistics courses by preparing 20 lectures worth of comprehensive R/Python code
• Gave an invited lecture on missing data handling and multiple imputation

• Improved the website engagement rate of young adult users by > 10% by designing and directing over 20 discovery interviews
• Contributed to the launch of a new mobile app by conducting >10 usability tests

Doctor Of Philosophy - Phd, Clinical Psychology, Quantitative Methods

Master'S Degree, Psychology

Bachelor Of Science - Bs, Psychology