Share product ideas and upvotes with our product team
The hability to export all the guidelines as they are organized, will be super usefull to manage in this way to interact with AI models, check, improve, etc.
On the monitors feature, it would be nice to filter conversations based on the user's recent page views registered on conversation details, based on the workflow, or other data available on the APIs integrations.
For specific filters in Monitor- we would like to have a "match" option. As in - we are trying to identify different users with the same company ID and we want this to be dynamic. Meaning- if we have different users from the same company ID from many different company ID's we want to make a flag for a trend so they can be pushed to an agent ASAP an not stuck in a conversation with FIN. (As Monitors grows to monitor teammate conversations as well, this would be super helpful for making sure agents are working together and can identify trends/bugs quickly)Context - if multiple users reach out about a bug from the same company, we want to have an opportunity to have shared communications with the users that includes them all in the communication/troubleshooting process. As opposed to the individuals not knowing that the other reached out about the same thing. This will decrease duplicate work done by our customers at the company who reach out to us, as well as our technical team having to interact with individual users separately creating duplicate work.Goal is to have one agent work with multiple users reporting similar/same issues to improve user experience.
Summary:Provide the ability to control Fin agent access to articles by default (either individually or a folder level default), and/or support for configuring this setting via API while creating/updating articles.Problem:Currently, all published articles default to having Fin access enabled. While we intentionally publish some articles as unlisted (accessible only via direct link), Fin can still surface these articles and information within to users.Our team frequently manages articles via API and is not always working directly in Intercom. As a result, we cannot reliably disable Fin access without manually logging in and updating each article.Why this matters:Unlisted articles are meant to be restricted to users with direct links Fin surfacing these articles can expose context-specific or limited-use information This can lead to: Customer confusion Incorrect guidance being shared Increased support friction Current limitations:No way to set Fin availability defaults for articles No API support to enable/disable Fin access No ability to control this at the folder level Fin audience settings exist at an individual article level, but are not configurable via APIProposed Solution Ideas:Allow setting a default Fin availability state (e.g., OFF by default) Add API support to control Fin access per article Enable folder-level settings to manage Fin access at scale Ensure unlisted articles can be excluded from Fin by defaultImpact:This would give teams better control over what Fin can surface, prevent unintended information exposure, and reduce manual workflows when managing content programmatically.
We have adopted FIN voice but when FIN is unable to handle the query it needs to be escalated to a human agent. It would be helpful if we can build an ‘AI Summary step into the workflows like we can with live-chat. This will streamline calls and prevent clients from having to repeat themselves.
In episode 13 of Support Stack, Thomas Hils (Head of Support @ Wingspan) shows how he’s rebuilt QA using Claude Cowork + Intercom, without adding another tool.It’s a really practical look at: why QA gets harder as Fin improves where AI falls down without structure how to actually make AI useful for QA Watch here:
I want to be able to use the API to search for conversations with a certain CX score within a particular date range. Can you please add CX score to Accepted-Fields? https://developers.intercom.com/docs/references/rest-api/api.intercom.io/conversations/searchconversations#section/Accepted-Fields
After learning in a Fin chat that functionality doesn’t exist, I asked it to submit a feature request, and was directed here. Customers shouldn’t have to take this extra step when you can automate it.
Currently you can attach OTP to Data Connectors. You can also I believe include them in workflows. However, you cannot use your built-in OTP within procedures even if your procedure utilizes a data connector that requires OTP. At least that is what your AI bot is telling me. It basically said I must bring my own OTP functionality to the table.We should be able to use the OTP you have within procedures.
Currently Fin will only respond to the person who opened the case. In many cases this as the case changes other people might be added to the case but Fin will only respond to the original individual who opened the case. In some cases this person isn't any longer responsible for the case but is still the only one who Fin will respond to.Fin should be able to reply all in the case or respond to the last respondent and not default back to the OG case opener.
Hello suggestion to improve: when ticket conversation is automatically converted to ticket we have an issue that subject/title filed all the time with “Ticket” and there is no description.There is difficult to find needed ticket when half of your ticket has same subject- and has no sense.Also form customer pint of view - opening ticketing portal and all your ticket are ticket:Dyou have good enaught feature AI autofill- filling subject and description based on what is happening during conversation great! but it is not working with workflows….
It’s great that I can upload Knowledge base article to Intercom via API, but I’d really like to be able to do that for Fin Guidance and Escalation guidance (and potentially for other Fin training areas) as well. Otherwise I’m going to end up doing copy-paste. which is error prone.
We have some Guidance not Enabled and I would like to only see what is Enabled. This is not possible and seems like a simple filter that is similar to ‘Live’ on workflows.
Hi all! Problem: Currently, Fin Procedures lack a native "Close Conversation" operator. When using a Procedure to escalate to an external CRM (like HubSpot) via a Data Connector, the Intercom conversation remains open. This results in double escalations because there is no way to end the Intercom session once the data has been successfully pushed to HubSpot.Why Workflows aren't a solution for us: I am specifically using Procedures because I require the Fin Conversation Summary to be pushed to HubSpot. Standard Workflows currently can't capture and map that Fin-generated summary.Current Workaround: I’ve had to build a secondary Data Connector that calls the Intercom API to close the conversation after the HubSpot push. This is redundant, and adds unnecessary complexity to what should be a native step.Requested Change: Add a "Close Conversation" action block within the Procedure builder (similar to how it exists in the Workflow builder). This would allow for a clean "Exit" state after an external handoff is confirmed.
Escalation Rules are a great addition to the software but currently lack the flexibility of Guidance. My biggest gripe: Fin is finnicky (no pun intended) in scenarios when I want it to explain why it is escalating before it actually does. Instead, the moment an Escalation Rule triggers, Fin immediately ignores any other guidance I had written in the rule and hands back off to the workflow, which is a bizarre experience if you’re an end user.A few ideas of how to improve:Create a new workflow trigger titled “If an Escalation Rule is triggered” (so I can then personalize the resulting workflow). Allow for conversations to auto-tag upon a certain Escalation Rule firing. (That way I can modify my existing workflow to check for those tags and offer personalized messaging according.) I’m sure there are other ideas that fellow Intercom customers may have as well.Again, big fan of Escalation Rules as a concept, but currently it just isn’t as flexible as I would like.
It would be very useful if individual conversations could be manually saved/tagged/favorited to make it easier to find and revisit them later. I know there is a way to have Fin ‘auto-tag’ conversations () but this is more of a manual process for times when you stumble upon something and want to check on it again later.
At the moment if you use Fin AI & have large responses from it it is pretty difficult to get rid of the keyboard when on mobile. It would be great if Intercom SDK had additional control over keyboard usage (e.g. hiding it on message sent to have full screen allocated for response).
Already have an account? Login
No account yet? Create an account
Enter your E-mail address. We'll send you an e-mail with instructions to reset your password.