Introduction to training Fin
Understand the Train section and the three essentials for shaping Fin's performance.
Training is how you turn Fin into a powerhouse for your business. It’s where you’ll spend a significant amount of time as you look to resolve more and more complex queries with Fin.
But you’re not starting from scratch. Fin is built on the Fin AI Engine, our patented architecture purpose-built for customer service and trained on millions of real support interactions. When you train Fin, you provide that final vital layer of business DNA: your unique content, your unique procedures, your unique brand voice.
Video transcript
Training is how you turn Fin into a powerhouse for your business, and it's where you'll spend a significant amount of time as you look to resolve more and more complex queries with Fin. But before we dive into the tools available to you, I want to reassure you that you're not starting from scratch. In fact, you're standing on the shoulders of some incredible engineering. We call it the Fin AI Engine. It's our patented architecture, purpose built for customer service, and trained on millions of real support interactions. Our engine works behind the scenes to refine every query, optimize every response, and validate the quality of every answer. Think of this engine as the foundation. When you train Fin, you provide that final vital layer of business DNA, your unique content, your unique procedures, your unique brand voice. Your investment here is the single biggest factor in Fin's performance for your context. So that's what the Train section is all about. Here, you'll have everything you need to configure that final piece of the Fin puzzle. But to start, you'll need to focus on a few critical sections. The first is Content. It's where Fin learns from your help center, docs, and websites. And without it, Fin can't know the actual answer to your customer's questions. The next is Guidance. Guidance acts as the employee handbook, defining your tone of voice and the specific logic Fin should use. It bridges the gap between what Fin knows and how Fin behaves. And lastly, Escalations. Escalations give Fin a playbook to know exactly when to hand off to your team, which is of course critical to building out your complete customer experience. As we said before, you'll be spending a lot of time here, especially in early development, but you don't need to configure everything at once. Start with the essentials we've just outlined. What's cool is that it's similar to training a new team member, but with one massive difference, scale. With your team, you have to train each agent individually, and then they go on to help one customer at a time. But when you train Fin to answer a question correctly, it applies that knowledge to future interactions immediately, whether that amounts to ten customers or ten thousand. So you're ready to get started. Your next step is to choose which content to give Fin.
The three essentials in Train
The Train section gives you everything you need to configure Fin for your business. To get started, focus on these three areas:
| Section | What it does | Why it matters |
|---|---|---|
| Content | Where Fin learns from your help center, docs, and websites. | Without it, Fin can’t know the actual answers to your customers’ questions. This is your starting point. |
| Guidance | Acts as Fin’s employee handbook, defining your tone of voice and the specific logic Fin should use. | It bridges the gap between what Fin knows and how Fin behaves. |
| Escalations | Gives Fin a playbook for when to hand off to your team. | Critical for building out your complete customer experience and ensuring the right conversations reach the right people. |
You don’t need to configure everything at once. Start with these essentials and expand from there.
Training at scale
Training Fin is similar to onboarding a new team member, but with one massive difference: scale.
With your team, you train each agent individually, and they go on to help one customer at a time. When you train Fin to answer a question correctly, it applies that knowledge to every future interaction immediately, whether that’s ten customers or ten thousand.
