Data Scientist Internship — CryptoMize

About the role

You will build clustering, classification, and time-series models on real engagement corpora, backtest them against outcomes we have already recorded at 89% historical accuracy, and learn what the textbook quietly omits — that data in the wild arrives dirty, adversarial, and mid-crisis.

What you'll do

• Build clustering, classification, and time-series models on real engagement corpora, backtest them against outcomes we have already recorded at 89% historical accuracy
• Run the full modeling loop on real corpora: cleaning, feature engineering, model selection, validation, and the written method note a strategist can audit
• Use: Python 3.11, Prophet for time-series forecasting, Jupyter Notebooks with versioned outputs, PostgreSQL for extraction

What you need

• Python
• Machine learning
• SQL
• PostgreSQL
• Any graduate