Customer chat answers from the help center with human handoff
An AI agent answers customer chats from help center content and account data within set limits, and hands off to staff when needed.
The use case
Support teams at growing companies get many chats about the same topics, like refund status, account changes, and how-to questions. An AI agent answers from the help center and, where permitted, looks up account or order details, resolving routine questions in the customer's language. Klarna reported its assistant handled a large share of chats in its first month. Staff still handle complaints, disputes, and anything outside policy, review samples of AI conversations, and fix help articles that cause wrong answers. Customers should always have a clear way to reach a person.
Example workflow
- Clean up help center articles
- Connect the agent with limited account lookups
- Set topics that always go to a human
- Launch on one channel and monitor
- Review samples and fix articles weekly
Tools
- Intercom Fin
- Zendesk AI agents
- Salesforce Agentforce
Risks and guardrails
- Wrong answers from outdated articles
- Customers trapped without a human
- Account actions need strict limits
- Review conversation samples
Sources
Written by Agents Lookup. Credits and official pages for the tools this use case names:
- Klarna press release on its AI assistant: www.klarna.com/international/press/klarna-ai-assistant-handles-two-thirds-of-customer-service-chats-in-its-first-month
- Intercom Fin: www.intercom.com/fin
- Intercom Fin (official page): www.intercom.com/fin (checked 2026-11-09)