NLP. Machine Learning. Analytics. Strategy.
Data science PhD researcher turning online behavior, text, and market signals into evidence people can act on.
Working pattern
Public community data, search signals, entertainment data, research sources, and stakeholder evidence.
Cleaning, schema mapping, deduplication, text preprocessing, feature construction, and scoring rules.
Models, rankings, robustness checks, error analysis, and concise business-facing explanation.
Selected analytical projects
PhD research project
Analyzes public online community discourse around product launches through collection workflows, text preprocessing, NLP feature extraction, sentiment analysis, and release-relative outcomes.
MSc thesis
Quantitative study of whether key personnel are associated with box office performance, using entertainment-industry data, statistical modeling, robustness checks, and interpretation.
Research assistant project
Repeatable evidence-monitoring workflow for emerging AI developments, turning academic and practitioner sources into structured summaries, research decks, and stakeholder briefings.
Portfolio project prototype
Movie performance tracking concept that collects weekly movie data, standardizes ranking inputs, and scores titles using a predetermined points function.
View repoIllustrative portfolio project
Ten-day simulation of the Movie League concept. Every second represents a new daily refresh and the table re-ranks.
| Rank | Title | Daily gross | Momentum | Audience | Reviews | Score |
|---|
Experience path
Warwick Business School
Builds release-relative data workflows connecting online discourse, search behavior, participation dynamics, sentiment, and market outcomes.
Technical stack
Used Python to scrape and clean movie/community data, engineer features, score rankings, and structure reproducible notebooks.
Next step
