W

Welton Wang

move fast. break things. buy duct tape.

About Me

Born and raised in San Jose, California. Amateur cooking enthusiast. Studying computer science (AI/ML and building scalable systems) at Stanford University. Interested in media, business, psychology, politics, and social behavior. Likes shooting pretty pictures, watching interesting movies, and taking long walks. Would like to own a dog one day.

Projects

things I've thought about

Moderate Hatespeech

Replacing Reddit Mods

ModerateHatespeech is a scalable, automated, community moderation provider that contextually flags, tracks, and removes abusive user-generated content. Backs some of the internet's largest communities. Built in PyTorch + Docker, running on RunPod.

250M+

Users managed

2B+

Comments analyzed

In collaboration with:

Better Food Recs

Because Yelp sucks

BetterFoodRecs aggregates public Reddit data on-demand, using pre-GPT era natural language processing to extract, analyze, and rank geographically-refined restaurant recommendations. Built in Python + Sanic + Transformers.

The Bipartisan Press

Politics, politics, politics

Formerly a national news outlet, The Bipartisan Press's Political Bias API used transformers to measure political bias in long-form content. Built with Tensorflow, Python, Uvicorn, and has powered hundreds of newsletters, aggregators, projects, and studies.

Contact Me

Email
hello [at] weltonwang.com
welton [at] stanford.edu