Data/ML/AI
Transfer Learning
Also written as Pretrained Model
Starting from a model someone else trained on a very large dataset and adapting it to your narrower problem, instead of training from scratch.
Think of it like
Hiring someone who already speaks the language and teaching them your industry, rather than teaching them to speak.
Junior or senior?
Now the default approach, so claiming it is unremarkable.
Senior sounds like
Can say what they froze, what they retrained, and how little of their own data they got away with.
Ask them
“How much of your own data did you actually need once you started from a pretrained model?”