What you will do
- Clean and structure product and user-event data into training pipelines
- Implement, evaluate, and compare recommendation approaches (collaborative filtering, content-based, hybrid)
- Serve the winning model behind an API our products can call
- Write up what worked, what did not, and why, for both engineers and clients
What we expect on day one
- Python plus practical pandas/numpy use on messy data
- Coursework or projects in machine learning; scikit-learn at minimum
- Ability to read a research paper or library docs and turn them into code
Bonus points, not filters
- Experience with PyTorch or TensorFlow beyond tutorials
- Any deployed model, hackathon or otherwise, you can walk us through