Projects

A website enabling automatic ranking of online media outlets using machine learning models (for now, only using Unitary.ai’s detoxify model). Once per day, the headlines from the frontpage of multiple online media outlets are scraped, and sent to machine learning models.

This was a short project I did with a pick-up team for a weekend-long Hackathon. The idea was to take the most liked tweets for a specific user, and use a Large Language Model to generate new tweets in this user’s style.

A simple browser extension that hides YouTube video length for user-selected channels. Unlike other similar browser extensions, this one also hides the video length in video thumbnails. You can easily install it from the Chrome Web Store.

DeepArchitect is a framework for automatically searching over computational graphs in arbitrary domains, designed with a focus on modularity, ease of use, reusability, and extensibility.

City-GAN uses Conditional Generative Adversarial Neural Networks to generate pictures of fake buildings, with the architectural characteristics of a specific city.

The Traffic Flow Mapper is a prototype tool for visualizing network traffic in 2D.

MDCGenPy is a synthetic dataset generator made specifically for testing clustering algorithms. It allows for incredible flexibility in generating data with specific shapes with a low effort.

The NTARC database is a collective effort of labeling and categorizing research made in the network traffic analysis field.

Multilingual embeddings can be used for any Natural Language Processing task which applies to multiple languages.