Experience

• 

  Spotify

  Nov 2020 - Dec 2021

  Data scientist

  •Processed and cleaned data from over 10,000 smartwatch users to prepare it for machine-learningmodels.•Developed a mood classification model achieving 92% accuracy and 90% F1-score in detecting useremotions.•Used a combination of SVM, Random Forest, and LSTM models to classify moods, improving accuracyby 15% over initial models.•Implemented a hybrid recommendation algorithm, resulting in a 20% increase in user engagement and25% higher user satisfaction.•Combined matrix factorization and content-based methods, improving recommendation quality by18% compared to initial algorithms.•Monitored the recommendation system's performance, identifying and implementing improvementsthat led to a 10% increase in recommendation accuracy. Show less

• 

  Memora health

  Aug 2023 - now

  Data scientist

  •Preprocessed over 5TB of data by handling missing values, removing outliers, transforming categoricalvariables into numerical representations, and normalizing continuous variables.•Utilized feature selection techniques such as Recursive Feature Elimination (RFE) and featureimportance analysis to identify the top 50 most significant predictors of patient health outcomes.•Addressed class imbalance by implementing the Synthetic Minority Over-sampling Technique(SMOTE), which increased recall by 15%.•Collaborated with a team of doctors and nurses to provide data-driven insights, which contributed to a20% reduction in hospital readmissions and a 15% decrease in complications among high-risk patients. Show less