A live demo, with real runs in it. Start free to connect your own →

demo These are the questions the eval asks and what a good answer has to contain. Editing one gives it a new identity, which is why an edited case stops being comparable to its own history — the reason this screen keeps versions rather than overwriting.

Test cases

8 cases in SupportBotQuality · v1

Recorded from a run

This is what your eval actually ran.

Taken from run 7CN4RVCP.

Ask A good answer contains
Can I get a refund without a receipt? receipt saved
Do you price match other retailers? no saved
Do you ship to Australia? no saved
What time do you close on a Tuesday? 5pm saved
user Hi, I bought a lamp last week
assistant Thanks for reaching out! How can I help with your lamp?
Can I still return it?
30 days saved
Is there a warranty on clearance items? no warranty saved
Can I change the delivery address after ordering? before dispatch saved
Do you offer student discount? no saved