Email is one of the most common support channels. It's also one of the hardest to get right with AI.
Unlike chat, email is much less predictable. Spam, auto-replies, and CCs need to be handled. Customers expect long, professional, exhaustive responses. PDFs need to be read and referenced. And escalations can get buried in a thread.
So that's what the team spent the last few months building for.
Fin now writes for the inbox – you can shape tone, length, greetings, and sign-offs so emails read the way customers expect. It handles email etiquette, managing CCs and reply-all with the rules you set. It reads and references PDF attachments. Fin follows up when a customer hasn't replied. And a new spam folder shows exactly why anything was filtered, with guidance you can write for what counts.
If you've tried Fin over email before and pulled back, this is the moment to try it again.
Watch the full walkthrough at the link below.
@Fin_ai is the future of customer support.
With the right knowledge base, it responds better than any human, is available on demand, and never spills coffee on its keyboard.
@intercom
AI isn't just for automation—it’s for insights. That messy spreadsheet, those endless convos, scattered docs? AI makes sense of it all. Fin can also do much more for customer support . Must-read from @brian_donohue on Intercom’s new CX score 👇
https://t.co/yGs6irmwDW
Our AI Group partnered with Messenger Team to challenge older design decisions through intensive experimentation. No stone left unturned! Join my demo to see how we use data to refine design at @intercom . #DesignThinking#DataDriven
New blog alert: we're now sharing our new ideas as much as possible!
We had a recent realisation inside Intercom. We’re having the most amazing conversations inside our four walls about building a cutting edge, valuable customer focused AI product, but if you didn’t work here, you’d never know.
We’re learning a lot fast, and we think we can help others learn too. So we’re doing a few new things:
1 We’ve launched a *new blog*, where we’ll share some of the latest ideas we have. Link in the comments!
2 We’ve started *building in public* more, sharing early demos of things we’re working on. We’ve shared 3 so far, and you can expect at least 1 a week going forward.
3 We’re also about to launch a dedicated technical AI blog, where our AI team will go deep on the experiments they are running, and what we’re learning.
Our goal with these is to hopefully help you learn from what we're doing, get you thinking, and also to start some discussions so we can learn also.
We've more to come, watch this space. Link in the comments to the blog where you can also watch our demo videos :)