Features

AI agents

Set up an assistant with its own name, tone, language and knowledge, then use it on any channel.

An AI agent is the part of DMLakes that holds a conversation: who it is, what it knows, which language it speaks and how it behaves. You build it once and reuse it wherever it is needed.

What it does

  • Describe how conversations should go in plain language: the tone, what to ask, what to avoid and when to stop.
  • Give an agent its own knowledge and tools, and keep several agents for different jobs, such as support, sales or an event.
  • Change an agent in one place and have the change reach every conversation it answers, on every channel.
  • Route a conversation to a different agent part-way through when the topic changes.

What to know first

  • Replies come from an AI model and can be wrong or out of date. You decide what the agent may talk about, and which situations go to a person.
  • The agent is only as good as the instructions and information you give it, which is why testing before going live matters.
  • You remain responsible for the messages sent from your account and for telling people when they are talking to an AI where the law requires it.

See it on your use case

A short demo shows how this would work for your own enquiries.

Book a demo