01 · MACHINE LEARNING
Travel Insurance Purchase Prediction
End-to-end classification with leakage-safe preprocessing, cross-validation, model comparison, tuning, interpretation and explicit failure modes.
View case study →01 / Featured
Selected data science & AI work
Five projects that show predictive modeling, statistical reasoning, computational research, applied AI and visual analytics—with limitations and role attribution made explicit.
02 · PUBLISHED RESEARCH
Computational Behavioral Research
KDD 2011 research applying data-mining and graph-based approaches to RFID-derived behavioral data and social organization.
View research case study →
03 · APPLIED AI / PRODUCT SYSTEMS
AI-Powered Translation
Product/system leadership across NLP, computer vision, multilingual translation and accessibility-oriented workflows.
View case study →
04 · SQL / STATISTICAL ANALYSIS
Mental Health in Tech
SQL-to-Python analysis of workplace mental-health survey data with prevalence estimates, confidence intervals and explicit sampling-bias analysis.
View case study →
05 · EDA / VISUAL STORYTELLING
IMDb Top 1,000
Exploratory analysis of acclaimed films across ratings, genres, votes, runtime and commercial performance, focused on clear visual storytelling.
View case study →02 / Technical work
Analysis that survives questions
Additional work demonstrates experimentation, inference, forecasting, optimization and decision support—not just notebook completion.
EXPERIMENTATION
Experimentation Lab
Retention testing and multi-arm promotion analysis with sample-ratio checks, confidence intervals and business interpretation.
Project archive →STATISTICAL MODELING
Red Wine Quality
Explanation-first regression with multicollinearity diagnostics, transformations, robust inference and a holdout check.
Project archive →MARKET ANALYTICS
Coursera Course Landscape
Ratings, enrollment, difficulty and provider concentration across a public course dataset, with limits of a static catalog snapshot made explicit.
Project archive →DECISION SIMULATION
Demand & Inventory Simulator
Forecasting and scenario analysis framed around a practical decision: how much to stock when demand is uncertain and spikes matter.
PREDICTIVE DECISION LAB
Customer Retention Decision Lab
Customer-risk modeling organized around intervention thresholds, error costs and ranking quality rather than a single accuracy number.
EVIDENCE-GROUNDED AI
Evidence-Grounded Research Assistant
A research workflow designed to keep generated answers attached to supporting passages and make uncertainty visible.
OPTIMIZATION
Swarm Logistics Simulator
A swarm-inspired optimization experiment framed around measurable objective improvement and repeatable scenarios.
03 / Background
A technical profile with unusual depth
Research, behavioral science and product leadership are not separate stories here—they shape how I frame data problems, test assumptions and translate analysis into decisions.
Scientific research
Graduate training in Neuroscience and Forensic Psychology, with quantitative behavioral research and a KDD 2011 publication applying data-mining techniques to RFID-derived animal behavior.
Research case study →AI & product leadership
8+ years leading digital products across AI/NLP, healthcare technology, telepsychiatry, fintech and SaaS—connecting user problems, technical constraints and delivery.
Applied AI case study →Behavioral & healthcare domain
20+ years of behavioral-health experience adds practical understanding of assessment, longitudinal observation, evidence quality, clinical workflows and real-world adoption.
How I work
Define the decision first. Build the simplest defensible analysis. Protect the evaluation. State failure modes. Separate evidence from inference. Make the result usable.