1. Knowledge Base
  2. AI Blocks and Flows

Should I build a custom AI Block model or use a template?

The two types of solution to your automation headaches

We have added templates to make your life easier. Depending on the issue you are trying to solve, you can use them to hit the ground running. You also have the option of building a custom model from scratch with little extra effort. Simply add data to your labels, that's it. 

Which option should you choose?


Do you have your own data?

The biggest factor you should consider is how much data you already have. For example, if you are classifying your own emails, chances are your inbox is stuffed full of perfect training data. This would be a great reason to choose a custom block.

If, however, you're trying to build a model for future use, such as for a clothing website where you need to categorize fashion inventory, our pre-trained Template will be the easiest choice.

How custom is the data you will be classifying?

No two email inboxes are alike. While some email responses will follow similar patterns (such as outbound marketing email responses), some industries receive specific content that our general Templates do not include within the standard training data.

In this case, you could either begin with one of our similar templates and add additional data to it and retrain, or you could build yours from scratch with your existing mail as training data.

Are you still exploring the capabilities of AI?

If you're still on the fence with how AI works, we recommend playing with our Templates first. Why? Because seeing a prediction happen is often the moment people realize the power of AI and have their "aha!" moment.


If you are getting stuck in the process, or want to discuss what solution is best for you, please feel free to reach out to us on the in-app chat.