Use Customer.io with AI
UpdatedAI is exciting and new—to you and us! Learn more about how we’re enabling AI features across our platform.
Check out our roadmap to see what’s coming next
We’re working on some new features—including an MCP server and an AI-assistant in Customer.io. Take a peek at our roadmap to see what’s in store for the future.
Our AI philosophy
We live in an incredible era, where you can ask a question or describe a thing you want to make happen, and AI just does it. We want Customer.io to work that way for you.
To make that dream possible, we’re working on a layer of tools that function across our platform, not just bits and pieces. We want your experiences using LLMs with Customer.io to be seamless, consistent, and reliable across the platform.
That takes time to build, but we’ll update this page as we introduce these features. You may also see notifications in our application that tell you when we roll new AI-based features out to your account.
Our current AI tools
We’re actively working on a tool that you can use across Customer.io, but in the meantime we support:
- Email content analysis: Use AI to analyze your email for better ways to deliver your message to your audience.
- Generating segments with AI: Give us a plain-text description of your segment and we’ll build it for you.
- In-app message suggestions: We’ll suggest some in-app messages for you based on your campaign workflow, goals, and audience, so it’s easier to craft messages that resonate with your audience.
- Using our documentation with AI: You can download plain-text versions of our docs including our API specifications to create your own GPT or get help with Customer.io from LLMs outside our platform.
Best practices: making sure AI works for you
When you create attributes, events, campaigns, segments, and so on, you give these assets names and descriptions. Our AI assistant relies on these names and descriptions to understand your data and provide AI-based suggestions—just like any member of your team would. That means that the more descriptive you are, the better the assistant’s suggestions will be.
For example, imagine that you have an attribute called cname
. This attribute might be relevant to you. But, if you don’t provide a description, the assistant might not know what it means. Is it a company name, customer name, or does it refer to the canonical name in a domain name system? That’s something you need to tell the assistant.
We don’t store or train AI on your customers’ personal information
We only use the names and descriptions of data (key names, not values) with AI. For example, if you provide a first_name
attribute for a person named “Alex,” our AI-based tools will know that you have a first_name
attribute, but will not see Alex’s name.
Follow uniform naming conventions for your data
When you create attributes and events, we’ll use the names you provide to understand your data and provide AI-based suggestions. To make sure that we understand your data—and produce suggestions that match your standard formatting rules—we recommend that you follow uniform naming conventions.
For example, you should follow a consistent naming scheme like snake_case
or camelCase
for your attributes and event data. That way, when we propose suggestions, they’ll match your preferred format.
Use tags to organize campaigns, segments, and more
Tags are a feature in Customer.io that help you organize similar assets like campaigns and segments. Tags not only help your team understand things in Customer.io that share common characteristics or use cases, they help AI as well!
Tags also have descriptions. Provide descriptions for your tags to help your team and our AI features understand your data!