Built a fully local research agent that explores any topic, synthesizes information, and generates summaries without relying on external APIs. The agent can search the web for context and works efficiently on Apple Silicon hardware. An experiment in privacy-preserving AI.

Developed and trained a classification model for predicting passenger satisfaction on high-speed rail. Performed exploratory data analysis, engineered features, and compared decision trees and random forests. Learned the importance of data quality and feature engineering.

A multi-agent research crew that leverages specialized roles to investigate topics in depth. Runs entirely locally with Ollama, demonstrating how agent architectures can distribute complex tasks effectively.

An AI-powered platform exploring how technology can support doctors and patients in managing health journeys more effectively. Focused on practical applications of AI in healthcare with a human-centered design approach.