We want to move a batch-scored model to online scoring. What changes, and what is point-in-time correctness?
Answer it out loud before you open anything. The value of the flags below is in comparing them to what you actually said — including whether you asked about the data before naming a model.
A bank scores every customer nightly for a pre-approved credit-line offer and shows the result in the app the next day. Product wants the offer to reflect a deposit made a minute ago, so they ask to "make the model real-time". The training set was built from a nightly warehouse table.
What it is really testing
Whether the candidate can turn "real-time" into concrete questions — freshness needed, latency budget, volume, which features change fast — and whether they understand that moving to online scoring changes the *features*, which means the training set must be rebuilt point-in-time or the online model will be evaluated against a world it never trained on.