Customer support · Production

Hospitable uses Intercom's Fin AI chatbot to resolve 30% of inbound queries and cut response times by 95%

The problem

Hospitable faced a growing backlog of support queries and was approaching the point of needing to expand headcount, while also lacking around-the-clock coverage for hosts who contact support outside business hours.

Workflow diagram · grounded in source
1
Customer submits support query
trigger
“a customer base of hosts who often need to contact support outside of business hours”
2
Fin searches knowledge base
ai_action
“Fin uses information from our comprehensive knowledge base to provide customers with solutions to commonly asked questions”
3
Fin resolves covered queries
output
“Fin is already resolving 30% of all inbound queries it interacts with, saving the Hospitable team valuable time”
4
Route uncovered queries to agent
routing
“If a question is not covered in our knowledge base, it routes the customer to one of our support agents so they can easily take over and provide further assistance”
5
Human agent takes over
human_review
“The handovers between Fin and our support team feel very natural and seamless for our customers”
Reported outcome

Fin instantly cleared the 500-query backlog, now resolves 30% of all inbound queries, and has reduced response times by 95%, while 61% of new customers prefer the faster AI responses over waiting for a human agent.

Reported metrics
inbound queries resolved by Fin30%
Backlog queries instantly resolved500
Response time reduction95%
new customers preferring AI responses61%
Reported stack
IntercomFin
Source
https://www.intercom.com/customers/hospitable
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

Fin instantly cleared the 500-query backlog, now resolves 30% of all inbound queries, and has reduced response times by 95%, while 61% of new customers prefer the faster AI responses over waiting for a human agent.

What tools did this team use?

Intercom, Fin.

What results were reported?

inbound queries resolved by Fin: 30%; Backlog queries instantly resolved: 500; Response time reduction: 95%; new customers preferring AI responses: 61% (source-reported, not independently verified).

How is this customer support AI workflow structured?

Customer submits support query → Fin searches knowledge base → Fin resolves covered queries → Route uncovered queries to agent → Human agent takes over.

Related customer support cases