Smart Eats is a lightweight AI engineering demo that simulates an Uber Eats-style food-recommendation system.
Save your time in finding target restaurant in your city!
In just a few clicks, users can enter their city, food craving, budget, and optional dietary restriction to get AI-generated Top-5 restaurant picks with explanations.
pip install -r requirements.txt
python -m src.data_prep
python -m src.embed_index
streamlit run app/app_streamlit.pyPlace the restaurant datasets (or demo Kaggle CSVs) inside the data/ folder.
The app runs fully locally; with an OpenAI API key, GPT will re-rank and explain recommendations.
End-to-end AI engineering pipeline: data prep → embedding → model re-ranking → web UI
GPT-based reasoning: natural-language explanations for each recommendation
Fallback mode: works offline using heuristic scoring when API quota = 0
Streamlit UI: clean interface for quick testing