Only if you want it to. The common deployment runs both: SleekAI handles the first reply with site context, LiveChat handles human escalation. SleekAI's display conditions decide which pages it owns, and the LiveChat widget can remain loaded elsewhere or be invoked via a preset when the user explicitly asks for a person.

Through a SleekAI preset that calls LiveChat's JS API to open the chat window with a prefilled opening message. The opening message includes the SleekAI transcript snippet relevant to the question, so the agent does not start from zero. Some teams also push the transcript into LiveChat's custom variables so it shows up in the agent panel automatically.

No. LiveChat keeps its transcripts in its own cloud. SleekAI reads from WordPress data and from its own conversation log. If you want SleekAI to learn from LiveChat history, export the relevant transcripts and feed the patterns into the system message or upload them into the OpenAI Files vector store. Once they exist as SleekAI context, they influence answers.

Only if both widgets are visible on the same page. Display conditions stop that. SleekAI typically owns marketing pages, docs, and product pages. LiveChat sits behind a button labelled "Talk to a human" inside SleekAI, or stays as the primary widget on the logged-in customer dashboard. The visitor sees one entry point per page.

Those still work on LiveChat-handled chats. SleekAI has its own logging and you can add a thumbs up or down via the JS API at the end of each session. Pair both and you get rating data for the human queue and for the automated tier, useful when arguing the bot's accuracy.

Yes, but the other way around is more common. SleekAI is always on, including out of hours, and LiveChat's office hours decide whether the human handoff goes live or queues as a ticket. If you want SleekAI to silence itself when an agent is online, a small custom check based on the LiveChat presence API can hide the SleekAI widget on those pages.

SleekAI data stays in WordPress: chatbot configurations, system messages, and conversation logs. LiveChat data stays in its own cloud. The handoff payload from SleekAI to LiveChat is a single string at conversation open, which the agent sees. Beyond that they do not share storage, so a SleekAI log delete does not affect LiveChat history.

Most teams use a fast cheap model like a Haiku or a small Gemini variant for first-reply pre-sales, and reserve a stronger model for the rare docs-heavy questions. Multibot lets you split that across two chatbots on the same install, so the volume of cheap questions stays cheap and the harder ones get the better model only when they hit.