Analyzed the global correlation between Corruption (CPI) and Human Development (HDI). Built a publication-ready, Economist-styled scatter plot using R to highlight governance trends and outliers.
Built a multi-step Deep Research Agent using Pydantic AI and GPT‑5‑mini that orchestrates DuckDuckGo web searches to generate structured, evidence-based reports on stock tickers and complex general topics.
Intent-aware stock vs. general research
Parallel deep dives with cited evidence
PythonPydantic AIGPT‑5‑miniDuckDuckGo search toolsStructured research reports
Built regression-based cost models in Python to quantify how age, BMI, number of children, smoking status, and region influence annual medical insurance charges and risk tiers.
Developed full-stack AI platform with HTML/CSS/JavaScript frontend and n8n-powered Agentic AI backend to deliver personalized CTE & dual-enrollment recommendations for DC students.
Full-Stack DevHTML/CSS/JSn8n WorkflowsAgentic AIAI for Good
Applied simple linear regression in R to analyze how global availability impacts Netflix movie ratings, revealing statistical patterns in worldwide content performance.
Analyzed 1M+ Spotify tracks in R exploring how danceability, energy, genre, duration, and temporal trends influence song popularity using correlation and regression models.
Building LLM-powered analytics infrastructure for the UN CRPD Dashboard. Synthesizing data from 190+ countries to track global treaty implementation ahead of the 2026 UN Conference.
Analyzed 228,711 records to identify churn drivers. Optimized detection of at-risk policyholders by 2x using LASSO and Random Forest models for revenue protection.
2x Risk Detection Gain
LASSO & Random Forest
R Languagescikit-learnChurn AnalyticsPredictive Modeling
Analyzed socioeconomic predictors of early death across 3,183 U.S. counties. Identified child poverty and income inequality as primary drivers using OLS regression.
3,183 County Dataset
Tableau Dashboards
OLS RegressionTableauPublic Health BIData Analytics