AI Agent
Help Centre articles for the Bookboost AI Agent, organised into six collections. Open a collection to see the articles inside it.
24 articles
Setting Up Your First AI Agent
Setup guide
- Draft mode: letting an operator check the agent before it sendsWhat draft mode does, how an operator reviews a drafted reply, and what it does not change.
- Setting a greeting and a sign-offThe two optional instructions that shape how your AI Agent opens and closes a conversation, when each one fires, and how they work alongside tone.
- Creating a new AI AgentThe end-to-end flow for creating a new AI Agent, from Messaging settings to the Knowledge tab.
- Choosing a starting templateThe four AI Agent templates, what each includes, and how to choose the right one.
- Naming your AI Agent and setting toneHow to name your AI Agent, pick its tone, and where the other General tab settings live.
- Setting assignment rules and channel filtersWhat assignment rules do and how to create one to control which conversations the AI Agent handles.
- Activating an AI AgentWhat to check before activating an AI Agent, and what to watch after it goes live.
Building the Knowledge Base
The AI Agent answers from a knowledge base you provide. Every entry is a knowledge item: a short topic, and the content that goes with it. These articles cover what to put in, how to write entries the agent can match, the three ways to import knowledge you already have, and how to keep a large knowledge base organised.
- Importing knowledge into your AI AgentThe three ways to bring knowledge into an AI Agent without typing it, from your website, from a document, or from another bot, and how to review what comes back.
- Organising knowledge with tagsHow to tag knowledge items so a large knowledge base stays navigable, which tags Bookboost suggests, and how the filter works.
- How the AI Agent learnsWhat a knowledge item is, how the AI Agent uses what you give it, and the two things it knows without being told.
- Writing effective knowledge entriesSix rules for knowledge items the AI Agent can match and answer well, plus what to avoid.
- What to cover in your AI Agent's knowledgeThe topics a hotel AI Agent usually needs to cover, two starter sets for common agent types, and how to check an import for gaps.
- Testing your AI Agent with the Preview panelHow to use the Preview panel and the Testing tab to simulate guest conversations and check your AI Agent before it goes live.
- Building out your AI Agent's knowledgeThe Knowledge tab step by step, importing, adding items by hand, tagging, and activating your changes.
- Answering reservation questions: how guest verification worksHow the AI Agent answers questions about a guest's own reservation, why verification is required, and which channels can verify.
- How the AI Agent recognises returning guestsHow guest context lets the AI Agent adapt its tone to a guest's history and preferences, and how guest privacy is protected.
Handover and Escalation
What happens when the AI Agent hands a conversation to a human
Monitoring and Managing
Reviewing how the AI Agent is performing, running multiple AI Agents for different scenarios, and controlling when the AI Agent is active.
- Reviewing AI Agent conversationsWhere to find AI Agent conversations in the Unified Inbox and what to review in the first weeks.
- Running multiple AI Agents for different scenariosWhen one AI Agent is enough, when to add a second, and how knowledge works across several agents now that you can copy it between them.
Limitations and Support
What the AI Agent does not do today, common issues and how to resolve them, and how to get help from the Bookboost team.
- What the AI Agent does not do todayAn honest list of what the AI Agent cannot do yet: channels, inputs, actions, and guest data.
- Common issues and how to resolve themThe most common AI Agent issues — escalations, stale answers, over-triggering — and how to fix each.
- Getting support for the AI AgentThe two AI Agent support channels, which to use when, and what to include in a bug report.