What it is
Fin is Intercom's AI support agent. It can run inside Intercom or on top of another helpdesk.
What you'd use it for
You would use Fin as your first line of support: it reads your help content, answers the customer, takes actions in connected systems, and hands off to a person when it cannot help. It is priced per resolved conversation rather than per seat.
Examples of use
A customer asks a question in chat. Fin reads the company's help content, finds the answer, and writes a grounded reply, so the customer hears back straight away.
A shopper asks for a refund. Fin acts through a connected back-end system to process it, taking the action rather than just describing how it would be done.
A support team already runs Zendesk and does not want to switch. Fin runs on top of that helpdesk, adding AI resolution without replacing the tool the team knows.
A stream of routine tickets comes in each day. Fin resolves them, and the team pays per resolved conversation rather than for every seat.
A conversation reaches a point Fin cannot help with. It hands the customer off to a human agent rather than guessing at an answer.
A small team already works in Intercom. Fin runs inside that setup as the first line of support, so AI resolution is added without a new platform to learn.
How it works
Fin runs on what Intercom calls the Fin AI Engine: it interprets the question, searches your support content and connected data, and writes a grounded answer, with support for tools and multi-step steps.
How it compares
Fin's distinctive trait among the autonomous agents is that it runs on other helpdesks too, including Zendesk and Salesforce, not only its own. Where Decagon and Sierra aim squarely at enterprise, Fin reaches further down into smaller teams.
What others say
Fin has a Gartner Peer Insights page of user reviews, but no Magic Quadrant or Forrester Wave position. Its headline resolution rate is a vendor figure.