Use this article to add Fin AI Agent to your Workflows, customize its handover behavior, configure channel-specific settings, troubleshoot conflicts with existing workflows, and preview Fin before going live. You can set up Fin to respond within a workflow, or configure workflows to pass conversations to Fin.

Go to Fin AI Agent > Workflows to get started. You can add Fin to an existing workflow, or create a new one. The best workflow triggers to use with Fin are:

To enable Fin on particular channels or for a specific audience, click on the workflow trigger settings (first block in the Workflows builder).

If you're creating a workflow for email, you can add audience rules using email predicates.

Then add Fin to your workflow by creating a path and selecting Let Fin handle. This inserts a Let Fin handle step which can then be configured to control the behavior and handover.

The Let Fin handle step triggers a response after a customer sends a message — Fin does not automatically respond to quick-reply button selections. To address this, add a message action after the button prompting the customer to type their question, or use canned replies on specific buttons to provide an immediate answer.

By using branching, you can create different Fin AI Agent experiences for different audiences.

For example, for paying customers, you might want to hand over to a teammate when Fin can't answer but for non-paying customers, you might want to direct them to your Help Center.

To customize Fin's behavior in a workflow, click on the Let Fin handle step in the workflow path.

For chat conversations, if you don't want Fin to hand over to a teammate, you can configure the step to not set any expectations for human support.

For chat conversations, you can ask Fin to gather more information from a customer before passing to your team:

Fin will encourage the user to share more information. The exact message it sends will be similar to "Sure! While I connect you, could you provide us with more details about your issue to help us find an answer faster?" This will be adapted to the context of the conversation and your workspace’s Fin personality settings.

By prompting the customer for more context before Fin hands over to a teammate, there is another opportunity to get more and/or better information than presented the first time around. The benefit to this is two-fold:

We've tested this feature extensively at Intercom and observed an increase in answer rate, confirmed resolutions, and CSAT.

At the end of the conversation, you can decide what happens if the customer hasn't responded after Fin answers:

If using Fin over email, you can configure what happens if the customer doesn't respond after Fin's email:

If you'd like Fin to follow-up, you can also specify the duration to wait before checking if the customer still needs help (in days or hours) and customize the follow-up message.

At the end of a customer's interaction with Fin, you can ask them to give a conversation rating by toggling these settings on:

Choose when to auto-close conversations by specifying a length of time Fin should wait before a customer is considered inactive. If you've enabled a follow-up, it will be sent 4 minutes after the last customer message. Once the follow-up is sent, the auto-close timer you set below will begin.

Then select if you want Fin to automatically close the conversation after customer becomes inactive:

You can also Customize what Fin says when closing the conversation to make sure the customer got what they needed and remind them they can still speak to your support team, even if the conversation is being closed.

Use the When Fin closes a conversation trigger to automatically run a workflow whenever Fin closes a conversation. Unlike the Teammate changes the state of a conversation trigger, this fires on Fin-initiated closures only — giving you precise control over post-resolution automation.

Fin escalates conversations based on natural language cues — such as a direct request for human support, negative feedback, or repeated replies without resolution. If the customer gives positive feedback, Fin closes the conversation and marks it as resolved. Ensure assignment settings are configured to automatically assign escalated conversations to the appropriate team or individual.

If the customer gives negative feedback (e.g. "No that didn't help"), they will follow a handover path that you will need to configure in the workflow.

For example, you could set expectations for when your team will reply or collect additional information that will help route the conversation to the right team. This helps make the transition as smooth as possible and can save teammates' time and effort when picking up the conversation.

You can also pass to a reusable workflow if you have a common triage or data collection process that you want it to follow. Use the Pass to a reusable workflow step to hand Fin over to another workflow.

You can add a Let Fin Handle step to a reusable workflow and trigger Fin directly from the Inbox. Fin requires a message from a customer after the workflow is triggered before it will generate a response.

Fin requires a message from the customer before it generates a response. To handle this, add a message action before the Let Fin handle step asking the customer to type or clarify their question, as shown below:

When you add Fin to a workflow, you'll be able to experience a conversation with Fin using your workflow.

First Save your new workflow, then click the Preview button at the top of the Workflows builder.

This will give you an interactive preview of the experience when talking to Fin. Simply start asking questions you would expect to find answers to in the support content you added.

You can also preview Fin from any website where you have the Messenger installed. Click Preview from a website and enter the address, then click Open website to load the Messenger with Fin on that website.

If you are using other Workflows alongside Fin AI Agent, there are 2 triggers where they could conflict with each other.

When both Fin's Simple deployment and workflows are enabled for the same audience, Simple deployment takes priority and only Fin will respond. To allow workflows to trigger, adjust audience targeting so they do not overlap, or disable Simple deployment. In certain scenarios, Fin may continue to respond even after a human agent takes over because of overlapping triggers like "Customer sends any message". To prevent this, adjust workflow configurations to ensure Fin does not re-engage when a teammate is active, and review triggers for unintended overlaps.

Workflow will be shown with reply buttons. Clicking on a reply button means entering the Workflow and Fin AI Agent will not fire.

If you leave the composer open (Let customer type), customers can still type in their question and Fin AI Agent will take over (Workflow will be skipped).

However, if you disable the composer at the start of the Workflow, Fin will be blocked.

If Fin is being blocked because the composer is disabled at the start of the workflow, here are some troubleshooting tips: