Delivery
Proof of concept
A small build that answers a specific question about feasibility — not a small version of the finished product.
A PoC exists to reduce uncertainty. "Can a model classify our support tickets accurately enough to be useful?" is a good PoC question. "Let's build a bit of the system and see" is not.
The common failure is the PoC that impresses on five hand-picked examples, gets approved, and collapses on real inputs. Guard against it by defining what "good enough" means numerically before building, and testing on messy real data rather than curated samples.
A PoC that concludes "this will not work" has done its job and saved you the build.