Chatbot

Create an AI chatbot that chats with your customers and answers from the knowledge base you upload and attach to it.

What is a chatbot agent?

A chatbot is a text-based AI agent that holds conversations with your customers and answers their questions from your own content. You upload documents into a knowledge base, attach it to the chatbot, and the agent responds with grounded, accurate answers drawn from exactly the content you selected, in your customers' languages.

Each chatbot answers only from the knowledge base attached to it. That means you can run several chatbots side by side, each scoped to a different set of documents.

Creating a chatbot agent

Chatbots are created entirely through the Lucy dashboard. No coding, no API keys. Just configure, attach your content, and connect it to a channel.

Create a new agent

In the dashboard, click Create Agent and select Chatbot. Give it a name that describes its purpose, like "Support Bot" or "Sales Assistant".

Upload your knowledge

Add the content your chatbot should answer from. Create or open a knowledge base and upload PDFs, web pages, or text files. Lucy automatically processes, chunks, and indexes everything so the agent can search it instantly.

Attach a knowledge base

In the agent's settings, select the knowledge base you want this chatbot to use. The agent answers strictly from the documents in the knowledge base you attach, so it stays accurate and on-brand instead of guessing.

Set the greeting and behavior

Write the first message customers see when they open the chat, and describe how the agent should behave: formal and professional, friendly and casual, or sales-oriented. You can add instructions like "always ask for an order number first" or "escalate to a human when unsure".

Connect tools (optional)

Enable tools so your chatbot can take actions during a conversation: look up orders, check account balances, create support tickets, or book appointments.

How answering works

When a customer sends a message, the chatbot searches the knowledge base attached to it, finds the most relevant passages, and writes a natural reply grounded in that content. This technique is called retrieval-augmented generation (RAG), and it keeps answers accurate and traceable to your own documents. Update the knowledge base at any time and the chatbot immediately answers from the latest version.

Deploying your chatbot

Once configured, go to Channels in the dashboard and choose where your chatbot goes live. You can add it to your website as a chat widget, connect it to WhatsApp or Telegram, or embed it on your WordPress site. One chatbot can serve several channels at once.

Connect a knowledge base, pick a channel, and your chatbot is live, answering customer questions from your own content around the clock.

PreviousVoice Agents NextOverview