A data science project that recommends Starbucks beverages based on user preferences like taste, caffeine, calories, protein, and milk type. It combines machine learning, feature engineering, and interactive input handling to deliver personalized drink suggestions—merging tech with a coffee shop experience.
- Python
- pandas, numpy, scikit-learn, KNN, OpenAI (gpt-4o)
- CLI-based user input
- Modular design with src/ folder structure
The raw dataset is pulled from Kaggle (https://www.kaggle.com/datasets/henryshan/starbucks)
Clone the repo:
git clone https://github.com/your-username/starbucks-recommender.git
cd starbucks-recommender
Populate the .env file with the your own API key like below
OPENAI_API_KEY=
python main.py