Scale Machine Learning Engineer interview experience

The setup sounded creative and very CV-flavored: you’d connect several stars conceptually to form a larger “target,” and then the task was to compute orientation. I had to complete it within a couple of weeks, and the project came with datasets and starter scaffolding, with the expectation that I was comfortable in Python and PyTorch.

After that, the process moved into a more hands-on ML coding evaluation. I was given a Colab notebook and asked to work with raw data—reading it, predicting, and evaluating models provided in the notebook. The star-detection theme carried through, and the work