--- title: "AI Concierge: Overview" description: "> AI Concierge is a guest-facing chat assistant that understands plain-English questions, searches your live catalog, and answers about your properties in your brand voice. --- ## What AI Concierge does AI Concierge adds a smart chat drawer to your…" url: "https://homerunner.io/docs/homerunner-ai-concierge-overview/" updated: "2026-08-31" category: Getting Started --- # AI Concierge: Overview > AI Concierge is a guest-facing chat assistant that understands plain-English questions, searches your live catalog, and answers about your properties in your brand voice. --- ## What AI Concierge does AI Concierge adds a smart chat drawer to your HomeRunner site. Guests type what they actually want ("a dog-friendly place near the beach for 6 people in July") and the Concierge interprets the request, searches your real availability and pricing, and replies with matching properties and a conversational answer. It also fields property-specific questions ("is there a hot tub?", "tell me about the kitchen") using your knowledge base and photo data. It works on top of your existing HomeRunner Core site. Nothing about your booking flow, PMS sync, or property pages changes. The Concierge sits alongside them and turns vague guest intent into the right listing. ## Who it is for AI Concierge works for **any catalog size**. Whether you manage a handful of listings or several thousand, the chat assistant is available and answers guest questions about your properties. For larger catalogs, semantic search also solves a real discovery problem: when guests cannot scan the whole catalog by eye, the Concierge surfaces the right matches from natural language. See [Pricing & Plans](/docs/ai-concierge-pricing-plans/) for how the add-on is priced. ## How it understands guests ### Natural-language search Guests describe what they want in their own words. The Concierge maps that to real filters (location, dates, guest count, bedrooms, amenities, vibe) and runs a live search against your catalog, then explains the matches it found. ### Concierge Handpick Some requests describe a vibe or a visual detail that no filter can express (an open floor plan, a bright and airy feel, a black headboard in the bedroom). For those, the Concierge reads your listing descriptions and photo data and hand-picks the homes that genuinely match, instead of quietly falling back to a plain location search and leaving the detail unanswered. See [Concierge Handpick](/docs/aise-concierge-handpick/). ### Multi-intent messages A single guest message often carries more than one request ("somewhere pet-friendly with a pool, and what's your cancellation policy?"). The Concierge decomposes multiple intents in one message and handles them together rather than forcing the guest to ask one thing at a time. ### Smart dead-end recovery When a search comes back thin or empty, the Concierge does not just say "no results." It progressively relaxes the least important filters and offers **recovery chips** the guest can tap: alternative dates when the same property is open a week earlier or later, relaxed size or guest-count constraints, or a way to reach a human. When party size alone is the blocker, it states the actual capacity ceiling in plain language, and, if Conversion Kit's Group Stays is turned on, adds a link that opens the wizard pre-filled with the guest's dates and party size so they can pair homes together. See [Group Stays](/docs/ai-concierge-group-stays/). The goal is to keep the guest moving toward a booking instead of bouncing. ### Smart replies Beyond finding the right properties, the Concierge adds small touches that make replies feel considered rather than mechanical: a one-sentence acknowledgement of the request grounded in a real fact about your destination, a plain statement of the dates it picked whenever a guest was vague about timing, and click-to-apply alternative-date chips when a looser window would open up better options. See [Smart Replies](/docs/aise-smart-replies/). ### Photo-aware answers The Concierge can describe specific spaces in a property because it draws on per-room data extracted from your real photo gallery. Ask about the outdoor area or the kitchen and it cites features the model actually observed, never a generic guess. See [Photo Analysis](/docs/ai-concierge-photo-analysis/). ### Property-level knowledge On individual listing pages, the Concierge can open with a property-specific greeting and answer detailed questions using that property's knowledge base (house rules, local tips, owner notes). See [Property Chat](/docs/ai-concierge-property-chat/). ### A guided Home tab Turn on the tabbed widget shell and the chat drawer opens on a branded Home screen instead of a blank transcript: a resume card that picks up a returning guest's last exchange, your site's Popular Questions as tappable prompts, and an entry point for the homes you have not published. Guests who have set up date watches through Conversion Kit's Availability Watchers also get a Watchers tab to check on and pause them. See [Home Tab](/docs/ai-concierge-home-tab/). ## When a guest needs a person Some conversations are better handed to your team. When a guest reaches a dead end or asks for help, someone on your team can step into the live chat and take over the conversation in real time, then hand it back to the Concierge when they are done. You can also get a Slack alert the moment a guest leaves their contact details or gets stuck, and step into the chat straight from Slack. See [Live Takeover](/docs/aise-live-takeover/) and [Slack Notifications](/docs/aise-slack-notifications/). ## What you control - **Brand voice**: a short description of how your team talks, applied to every reply. - **Knowledge base**: per-property facts the Concierge draws on. - **Recovery and outreach**: how dead-ends are handled, including [lead capture](/docs/ai-concierge-lead-capture/) when a guest leaves their details. - **The chat drawer**: position, greeting, and on-page nudges. - **The ask bar**: a search-style bar you can place in your hero or search area that opens straight into the Concierge. See [Ask Bar](/docs/ai-concierge-ask-bar/). All of this lives under **AI Concierge** in your WordPress dashboard. See [Operator Tools](/docs/ai-concierge-operator-tools/) for the full admin tour. ## Measuring the value The **Performance** tab gives you a live view of what the Concierge is delivering: cost trajectory, booking attribution, response times, and recovery-chip engagement. It is the dashboard you would show a partner at renewal time. See [Performance](/docs/ai-concierge-performance/). If you run Google Tag Manager, the Concierge also pushes chat-open and results events to your data layer, so assistant usage folds into the analytics setup you already have. See [Google Tag Manager](/docs/aise-google-tag-manager/). ## Getting started Setup runs about 20 to 30 minutes from install to live. Walk through it in [Getting Started](/docs/ai-concierge-getting-started/), and review data handling in [Privacy, Data, and Cost Controls](/docs/ai-concierge-privacy-data-and-cost-controls/).