Channels
A channel is the medium through which users reach an AI agent: the surface where a conversation actually happens. A web chat widget, an API integration or a messaging app are all channels. The channel doesn't change what the agent knows; it changes where and how people talk to it.
Think of a channel as the doorway into the same room. Whether someone enters through a website, an app or an API, they arrive at the same agent, but the doorway shapes the experience on the way in.
Why channels matter
Separating the agent from the channel brings real flexibility:
- Reach: the same agent can meet users wherever they already are, rather than forcing them to one place.
- Consistency: one configuration can be delivered across several surfaces, so the answers stay the same everywhere.
- Fit: each channel can suit a different context, an embedded widget for your site, an API for another system, and so on.
Channels and touchpoints
A channel rarely stands alone. It is one of the ingredients bound together by a touchpoint, alongside the model, knowledge and tools. The touchpoint decides what the agent is; the channel decides how it is delivered. Choosing the channel is therefore part of configuring a touchpoint for its intended audience.
Channels in Rational AI
When you set up a touchpoint, you choose the channel through which it is exposed, so the same carefully configured agent can be served on the surface that suits each use case.
Additional resources
- Touchpoints settings: choose how an agent is delivered.
- Channels (Chat): one surface where users interact with an agent.