Data Science/Analyst Intern | Goodwill Industries - Charlotte, NC | 01/2025 - Present
- Created a database in SQL to concentrate data from multiple sources into a single database for the department.
- Goal: Enhance effectiveness of Goodwill Training Programs to increase postgraduate employability.
- Questions:
- How can we maximize skill confidence in those enrolled in Goodwill Training programs from feedback in surveys?
- What features separate those able to find employment after graduation from Goodwill University from those who aren’t?
- Can we use trainee personal and survey data to isolate problem areas within the program and begin to aid them?
- Utilized SAS and Python libraries (K-means, Linear Reg. ) to identify and minimize hindering factors from trainee data.
- Increased r by 2% (from 75% to 77%) by refining feature selection and reducing noise in trainee data.
- Used tools such as Power BI to create visual reports to investors for GoodWill University programs, expecting detailed insight on program success.
Mental Health App Development (Remote Internship) | Howard University-BRSP - Charlotte, NC | Data Science Researcher | 05/2023 - 09/2025
- Goal: Use smart watch health data and user anecdotes data to inform those with mental disorders about themselves and possible triggers for prototype.
- Questions:
- To what extent do the combination of physiological indicators explain/align with proclaimed psychological disorders?
- Can we use user personal data to identify triggers of these behaviors, and possibly predict another event?
- Conducted in-depth research analyzing mental health data to support app development using data manipulation and analysis with Python.
- Created database using SQL and analyzed data using Python libraries, such as pandas and NumPy. Built and evaluated predictive & classifier models (e.g., linear regression, decision trees) using scikit-learn to identify key variables for app prototype.
- Predicted the general usefulness of variables and improved R² score by 20% (from 0.45 to 0.65) by optimizing feature engineering for mental health predictions.
