Train Fin

We are just getting started on FinAI. When nobody is marked as available, but a user selects “talk to a person”, we’d like to train it to tell the user that nobody is here right now and we’ll get back to them soon.  I have given FinAI guidance in the Communication Guidance and Handover & Escalation sections, but it doesn’t give the answer I trained it to give, and our users are left hanging with the expectation that a live agent will be with them soon.  How do I train it to give specific information to a user to let them know we aren’t available right now and to leave their info?  Or is there a way to automatically turn off “talk to a person” if nobody is marked available?

Hey Community,We have been using FinAI for a while now and it has been great so far.I need this little adjustment but I cant seem to find my way through. I wouldn’t want Fin to say this ‘I have ingested your content, you can give me a spin! Just type any question your customers might have. Don't be shy, I love a good chat!’I cant seem to find how to turn this off.

I will have data of my customers on my website (e.g. hours played). I want FinBot to answer certain questions differently depending on this data. Is there any way to do this? 

I added a bunch of data from my database, but fin can’t answer a single question related to that.

My company has three products. Each of them have their own knowledge center. Each customer conversation is tagged based on which app the user originated from. Fin doesn’t seem to be able to connect the dots that when a conversation is tagged with Product A, it should use the associated knowledge center. I have provided this instruction via the guidance section of the training system.Do you have suggestions about how I can get FIN to focus it’s answers based on a tag?

Currently when syncing with Confluence, the only visible sources are those contained within ‘Pages’. Anything that’s in a folder or subfolder is hidden and so cannot be located and used for FIN or CoPilot. Is there a workaround for this that doesn’t mean we have to remove everything from the folders and have 1000 individual pages for our knowledge base? Has anyone found a workaround that isn’t too time consuming?

We have a large database of typical customer questions and answers. These come from a service desk that is not currently supported by Intercom. What is the best strategy to add this content to Fin Knowledge?There are too many entries to manually input them as snippets.We’re considering adding them as a PDFs, but we’re not sure how Fin will chunk the data—whether it might split Q&As incorrectly or merge unrelated entries.Is there any guidance on how to format the PDF, or an alternative way to upload the data in a Q&A format?

Is there a way to train FIN for more scenario based questions. Our customers have different use cases for our product and it would be very helpful if we can train fin to help answer scenario based questions. 

Hello. In the AI & Automation > Fin AI Agent > Setup > Content the limit for knowledge sources is 10. Aside from manually copying and pasting each link, is there a way to combine multiple knowledge sources with the same domain name into one? 

I’ve been testing Fin and I haven’t been satisfied with the responses just yet to turn it on for all users. I understand it’s based on our articles and related knowledge base, but I’ve found Fin to provide wrong or sometimes unhelpful information. Here are some examples:It frequently tells users to contact support to fix the issue (Isn’t this what people think they’re doing?) It get company branding incorrect. For example I work for a company called “G.J. Gardner” and Fin often defaults to “GJ Gardner”, missing the periods in the name.  Users often don’t provide enough info such as the record name in the software they want us to troubleshoot (i.e. a customer name or project number). How can I train Fin to inquire about this information? Fin defaults to US English. While this isn’t always bad, we have some industry specific terms that differ between contries (i.e “variation” in Australia and “change order” in the United States). How can I train Fin to use the correct terminology when re

I have a CSV list of 5000+ frequently asked sales tax questions and their corresponding answers. I would like intercom to ingest these questions surface the answer to customers. I would also like to be able to update these questions quickly over time as the answer changes. What is the easiest to ingest these into intercom? Should I use something like Confluence instead of CSV? Any advice? 

How I can get FIN to include information in its knowledge sources based on articles with Target Audience rules turned on?The scenario is that we have some articles in our HelpCenter that we want to restrict to certain groups of users. However when check the FIN knowledge base, none of the articles that I have Target Audience rules on appear here. Is there a way to add them to the knowledge base?

I tried to import external content in the Fin AI Agent--Overview interface, but it has not been completed since yesterday, and the prompt "Importing content, please wait..."The URL provided is the top-level domain The URL is public 

Before Fin AI starts drafting their response I want to give a short text input likeThis user is on the {plan} price plan with the following quotas: x seats, y integrations, ...they are using custom features: {a}, {b}, {c}Relevant metrics- metric_1: value_1- metric_2: value_2- ... so that the chatbot can give a much more contextual answer. This feels like low hanging AI fruit to get a much better answer. (ideally it also looks at some whitelisted custom data attributes (CDAs) / user tags)As far as I know this functionality does not exist.My proposal is to inject this via the javascript ( https://developers.intercom.com/installing-intercom/web/methods/#intercomboot-intercomsettings ) or via the backend REST API (for example via CDAs)Alternatively:  I could define a private help article that is scoped only to this person? (via https://www.intercom.com/help/en/articles/2982784-control-who-can-see-your-articles ).  But this really feels like a hack and not sure if it will be picked up by fu

Hello! Despite going over all the settings as noted in some of the existing FAQs such as: https://www.intercom.com/help/en/articles/8124534-how-to-use-pdf-file-content-with-fin, I still can’t get FIN AI to search through or pull answers from a PDF file I have uploaded. The upload went smoothly and within minutes was deemed good to go by the system (and it’s just a short document.)  Anyone have any idea why this isn’t getting going? FIN does pull from a couple of URL links we provided, just not a PDF.  Thanks in advance.

What are the most effective methods or tools for training FIN to conduct context-aware or context-driven conversational AI? Is it Workflow or a combination of snippets and articles? Is there a resource one can read to perfect the interplay between these tools?Is it possible to have FIN adjust its responses based on the user queries and preferences, providing a more customized and contextually relevant experience for the user?Example scenario:User: What are the best attractions in Paris?AI: The Eiffel Tower, the Louvre Museum, and Notre Dame Cathedral are popular choices.User: Can you recommend a good place to eat near the Louvre?AI: Certainly, there are several restaurants near the Louvre. Would you prefer French cuisine or something else?User: I would like to try authentic French cuisine.AI: In that case, you might enjoy 'Le Café des Musées' or 'Les Fines Gueules', both of which offer authentic French dishes and are within walking distance of the Louvre Museum.

How do we turn off the chatbot content from auto syncing?  It appears to be doing it every week.The situation for us is that we temporarily set our confluence wiki to public for initial import, to get thousands of content into the system, then we set it back to private. This works great to get all our content into the chatbot. But when it attempts to resync a week later, it all fails and all the content is now rendered useless. I’m sure their must be a setting somewhere to turn off each managed source from auto syncing, or at least change the frequency to something like 1 year.  But we can’t seem to find the setting, and FIN doesn't seem to know.  Perhaps maybe it has to be done on the back end. It seems like the most basic use case to just not sync imported content, so I’m hoping there is a toggle somewhere that I’m missing.  If not, then we’re pretty blocked from going to production and upgrading to the pro plan, as it takes hours to reimport the content, and we can’t do that over an

Hi, We have a website https://www.example.com, we want Fin to import contents from it, but exclude all contents in the blogs, with the following pattern: https://www.example.com/blogs/xxx How to do so?  

I was wondering, how exactly FIN using PDF files as a source of data. Is it the same as help portal where FIN reads, interpret the data and use it to compile answers or can it be used as Snippet? Example.Can I create a PDF file where I will list info in the following format If the customer asks “xxxx” question or its variationThen use this as a base for an answer - “yyy”I basically want to eliminate snippets with it. Use PDF file to gather all the snippets, then translate each PDF file to multiple languages and upload them altogether. Covering 2 points - managing multilingual content and uploading snippets-like instructions in a bulk. If Intercom team can take a look at it and let me know if it will work, that would be swell. And of course, if someone else tried this, plesae, let me know :)

Hi there!We have specific terms and languages that we use for the features in our product. When clients reach out to ask questions, they use different terms (i.e. memberships over plans) sometimes, which leads to Fin being unable to resolve their questions since the language in our content doesn’t match their language.Will Fin be able to pick up on these trends and build it’s own glossary of words that equate with the terms that the clients are using overtime? I would like to avoid changing the language of our features in our help center. I understand that there’s an option for things like snippets as well, but I’m wondering if there are features within Fin that could support with this.

Which format work best for Fin? It seems that Snippets are better processed by Fin. In this case, I am wondering if there is a solution to manage this content with efficiency. Is there a way I can isolate snippets on the same subject to make the changes needed? Or should I only count on my classification (titles) to do so? Is there a hierarchy in Fin’s choice to use an article or a snippet? What makes it choose an article from another? Or an article rather than a snippet? Is it based on the same algorythm than the one for the helpcenter?  Question: comment créer une facture?Answer based on those articles:Gestion de trésorerie - Comment importer manuellement une liste de factures ?Gestion de trésorerie - Comment supprimer une facture ?Gestion de trésorerie - Comment supprimer une échéance de paiement ?Utiliser les factures d'avoirWheareas I have an article called that does not appear at all Créer une facture, un devis, un document de vente 

I queried Fin to answer a question from one of our blog posts. While Fin did pull from our blog posts, he pulled an inactive link that resulted in a 404 error. I found the correct link but I don’t know how to correct in Fin. Can someone advise, please?For example: Say Fin pulled www.company.com/blog123However, that resulted in the 404 error. Fin should have pulled www.company.com/blogabcHow do I update the link in Fin? Thank you in advance! 

The external content parser is quite bad on FIN. Sometimes the webpage has a lot of content and the ‘preview’ is just a couple setences.If you have tables or other strucuture components, the parsing is really bad. 

Why is that that the quality of the answers is better - with the same questions- in them demo than in live Fin? I do not get why there is a difference and why the (paid) version available in our account does not answer as good as the demo on the website

Hi, Is it possible to add multiple snippets/records to FIN? Or upload JSON/CSV file?  Thank you,Noam