Conversation Starter: A Discussion on Our Revolutionary New AI Chatbot

The world of AI is changing rapidly, and at Intercom, we help set that pace. Today, we're excited to introduce Fin, our new chatbot powered by OpenAI's GPT-4 and Intercom's proprietary machine learning technology.

Just a few weeks ago we announced our first GPT-powered features in Intercom - a range of useful tools for customer service reps - and our customers have really appreciated the extra efficiency these features provide. .

The big goal, however, was to create a GPT-powered chatbot that could answer customer questions directly. To do this, it was necessary to be able to exploit the power of large language models but without the disadvantages posed by “hallucinations”. Initially, we weren't sure how long it would take to fix this, but now, with the release of GPT-4 by OpenAI, we can reveal that we've built a chatbot that can reliably answer questions customers at a high level. . We called it End.

In today's episode of the Inside Intercom podcast, I sat down with our Director of Machine Learning, Fergal Reid, to discuss our new AI chatbot, how we built it, what it does and what the next steps look like for that remarkable breakthrough.

Here are some of the key takeaways:

Our new AI chatbot can converse naturally using the latest GPT technology. Fin ingests information from your existing Help Center and uses only that knowledge, giving you control over how it answers questions about your business. At Intercom, we believe the future of support is a mix of bots and humans. Fin will not be able to answer all customer questions, and in these situations he can escalate more difficult questions to human support teams seamlessly. We've reduced hallucinations by about 10x, by creating constraints that limit Fin to queries about your business, based on a knowledge base you trust.

If you like our discussion, check out other episodes of our podcast. You can follow on Apple Podcasts, Spotify, YouTube or grab the RSS feed in your player of choice. The following is a slightly edited transcript of the episode.

A bot by any other name

Des Traynor: Welcome to an exciting episode of the Intercom Podcast. I'm joined again by Fergal Reid, our Director of Machine Learning, and he's going to talk about the launch of something we've been asked about almost every day since the launch of ChatGPT.

"This will actually be a bot that you can use for your business that has the natural language processing capability of ChatGPT, but will answer questions specific to your business"

Fergal Reid: Yes, thank you Des. Ever since ChatGPT came out, people were like, "Hey, can I use this to answer my support questions for my business?" And we were always like, "Oh, we don't know." We are not sure about the hallucinations. But today, I think we're really excited to announce that we're really excited about this product because we think we've done it. We think we've built something - it's actually going to be a bot you can use for your business that has the natural language processing capability of ChatGPT but will answer questions specifically about your business and we've built it using your help center so it won't answer random questions on the internet or anything. You can control what he says. The accuracy rate has increased significantly. We managed to significantly increase the accuracy rate thanks to the use of the new GPT-4 model from OpenAI and which you have access to in beta version. So I'm really excited about that.

Des: So the idea is that what people have been through and kind of fell in love with in ChatGPT, which is actually this bot that you can ask anything and give good luck respond. Can you do this for your business?

Fergal: Yes. Kind of. So we deliberately made it so that you can't ask him for anything. The idea is to build...