An agent must decide whether to resolve a Case automatically or escalate it to a human. Which design principle BEST supports this behavior?

The correct answer is B. Human-in-the-loop design. Human-in-the-loop (HITL) design explicitly builds decision points into an agent workflow where the AI determines whether it can resolve an issue autonomously or whether a human must intervene. This is the foundational principle for escalation logic. Hard-coded escalation rules…

AI Features for Service (e.g., Einstein Bots, Next Best Action)

Question

Options

AHard-coded escalation rules
BHuman-in-the-loop design
CAnonymous execution
DStatic response templates

How the community answered

(31 responses)

A
6% (2)
B
87% (27)
C
3% (1)
D
3% (1)

Explanation

Human-in-the-loop (HITL) design explicitly builds decision points into an agent workflow where the AI determines whether it can resolve an issue autonomously or whether a human must intervene. This is the foundational principle for escalation logic. Hard-coded escalation rules (A) are rigid and cannot adapt to nuanced case complexity. Anonymous execution (C) refers to running without user identity context and is a security concern, not a decision-making design. Static response templates (D) produce fixed outputs and cannot make dynamic resolution-vs-escalation decisions.

Topics

#Human-in-the-loop#AI design principles#Service automation#Escalation

Community Discussion

No community discussion yet for this question.

Full AI-201 Practice