Projects

All projects

2 projects

Cross-validation accuracy comparing four Titanic classifiers with simple and engineered features against a majority-class baseline.

Machine Learning

Titanic Survival Classification

A higher cross-validation score did not lead to better test predictions. I compared four classifiers on Titanic passenger data to examine how feature engineering and tuning affected survival predictions.

Python
scikit-learn
Pandas
+3

n8n workflow connecting Google Drive ingestion, Gemini embeddings, Supabase retrieval, and a chat agent.

AI Automation

RAG Document Q&A Agent with n8n, Supabase and Gemini

A document Q&A prototype built with n8n, Gemini, and Supabase. I restored the ingestion and retrieval workflow, migrated the embedding setup, and checked the answers against an OECD report.

n8n
Supabase
Google Gemini
+1