Share product ideas and upvotes with our product team
It would be fantastic if we could ensure that surveys are not sent in certain situations (such as if the customer contacted us about a non Support reason). Fin could analyze the convo just as it does for categorization and notation and decide whether or not a survey will go out based on parameters we have given it in Guidance. @Gerald Prado
Would be good to have other options not only close after some time but also assign to other team if customer is inactive for some time. Customer inactivity workflow is useless in this case and it is already reported via messenger
When Copilot suggest macros in the composer it doesn’t show the name of the suggested macro. Could it be possible that it shows the name of the macro being suggested?
I’d like to be able to filter my content suggestions by the team inbox it’s pulled the suggestion from. This is especially helpful for us, since we encounter very different types of enquiries, which we’d like to review separately. /automation/fin-ai-agent/content-suggestions
Hi There,Can I offer something to take Finn to the next level, and really get your cusotmers excited!I think voice command must be the future of live-chat. If the Finn widget could include the option to offer a ‘Voice’ response as an option. For example the visitor could tap a speaker button to ask question by voice, and get audio responses.For example, you could use an integration with https://elevenlabs.io/app/speech-synthesis/text-to-speech to give Intercom customer the option to set up ‘Cloned Voice’Then you could even offer website visitors the opportunity to converse with the voice of the Founder, or creator of the companies products, who could talk virtually with customers just as if they were talking directly to the real person.I feel this could really help conversion on your customers websites, as trust and customer service is everything, and this could potentially increase your overall business. Here’s hoping!CheersBrett
Hey Intercom Community! 👋🏄♀️Imagine being able to have natural conversations with Fin about your team's performance data right within Intercom. Instead of navigating through different reports, you could simply ask questions like:"How did Team A and B's response times compare last quarter?" "What patterns do you notice in our peak support hours?" "Show me our top performers based on CSAT scores"The magic would be in the conversation - just chat with Fin and get instant insights about:Teammate Performance Reports Response time trends CSAT patterns Workload distribution Custom metrics that matter to your teamThis would make data analysis more accessible to team leads and managers who want quick insights without diving deep into reports or exporting data.Has anyone else been wishing for something similar?
When I add a new item with the Knowledge tool, I should be able to apply the audience filters then and there instead of having to find the newly added item, click it, and then access the audience filter.
Requirement - 'I want to': Analyse conversation content in bulk using AI integration to identify themes and trends.End Result - 'So that': We can improve our help centre for Fin and make product enhancements through our ongoing product development.Action - 'When I do this': Access a function similar to Kraftful within Intercom, without needing to work with an API.Result - 'This Happens': The system automatically identifies trends and themes in conversations, eliminating the need for manual tagging by CS agents, which is often inaccurate or incomplete.
With multilingual Fin, users can engage with the bot in multiple languages. However, when asking to talk to a support rep speaking a different language, it’d be beneficial for that support rep to quickly gain an understanding of what has been discussed before getting back to the user.
Although FIN supports Dutch as a language, you are not able translate a text in a conversation into dutch using the AI tool. It seems more like it has been forgotten, so please add Dutch here. Thanks!
We’re a platform for collecting and reporting ESG data. While the platform interface is only available in English, we have a global user base. To make support as accessible as possible, I activated multilingual Fin in the most common languages spoken by our users. However, our help center materials are still only available in English due to resource constraints.Since the platform interface is in English, it’s crucial that users are guided using the original terms for sections, buttons, and actions. For instance, if our article on inviting colleagues says, “Go to the Configurations page and click Add colleagues,” it’s important that Fin doesn’t translate those terms. Currently, if a Swedish user interacts with Fin, they might be instructed to navigate to “Inställningar” and click “Bjud in kollegor,” which could cause confusion. It would be incredibly helpful if we could flag specific words or phrases as non-translatable within Fin (e.g., when writing help articles).
We have implemented multilingual Fin to support our users in 13 different languages. We’ve customized the welcome message for each language, but cannot customize the message sent when Fin closes the conversation or hands over to a support rep. For customizing the handover process to a human support rep, there are two main advantages:Managing language expectations:We’ve enabled multilingual Fin for 13 commonly spoken languages among our users. However, our support team doesn’t cover all these languages fluently. For instance:We've added Hindi support but don’t have Hindi-speaking reps. It would be beneficial to inform users during handover that our responses will be in English. For German, while we don’t have dedicated support reps, our German-speaking Customer Success Managers can assist depending on availability. In such cases, the handover message could inform users that they’ll hear back in either English or German.Setting expectations during holidays or other high-demand periods:With the holiday season approaching, response times may be longer due to limited staffing. The ability to customize the handover message would allow us to set clear expectations, such as specifying expected delays or providing alternative resources for immediate help.
It would be great if an internal notes section/ description section was added to Custom Answers. 😊This would allow you to add links to internal resources and give context to why this custom answer was created, especially if it is issue specific/ temporary and regarding something that will be improved in the near future.In this case: you could link the internal issue/ improvement URL in the notes/ description section and note down the date that you intend to keep this custom answer live and why etc. so that there is full transparency internally for the purpose of each custom answer.
When utilizing Fin over email, the responses and the pages users are directed to when selecting "That helped" or "Connect to an agent" are heavily centered on Intercom branding (see pictures below). It would be highly beneficial to have the ability to customize these pages and responses. Specifically, we would like to edit the colors and modify the content to better align with our product's branding. Our goal is to ensure our clients feel connected to our product, rather than to Intercom. We hope this feature can be considered for implementation.
When enabling the FinAI agent over emails, we cannot customise the webpage "talk to a person" when it’s clicked by the user. It would be very helpful if we could customise this webpage.
If there is a relevant product tour to answer a customer question, Fin should be able to automatically offer it to the customer within the chat to complete.
The large list of conversations is very difficult to scan because I can only read the (company) name of the person and the last message (which is often not descriptive enough). I would be useful if a conversation would be automatically titled by AI. Similarly to how ChatGPT chats are titled.
It would be great to have an option like 'Ask Another Question' to encourage users to keep engaging with FIN and explore more on their own.Right now, after FIN answers a question, users see 'That helped' or 'Wait for the team,' which either closes the conversation or puts them on hold for human support. This isn't ideal if we want to guide users toward self-serving. Adding an option that nudges them to keep asking could make a real difference."
I know you can train a custom answer to recognise queries by providing multiple different phrases and linking them to queries asked by customers. However when a bot is triggered by a query, it would be great for us to be able to mark that the custom answer was incorrectly triggered.Basically get it to learn from its mistakes so that if a similar query was asked again, FIN would know NOT to send that custom answer.
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.