Each assistant record has a remaining hours field stored in postmeta, updated when a new client signs or when monthly hours roll over. SleekAI reads that field at request time, so the bot only surfaces assistants with capacity left. Assistants at zero hours are filtered out before suggestions are generated.

Yes. Skill tags, languages, and time zone are all separate fields on the assistant CPT, and the bot can filter on all of them in the same query. A prospect asking for a French speaking bookkeeping VA in European hours gets matched against assistants who satisfy all three, not a fuzzy text search.

Custom briefs that don't fit the standard hour blocks are flagged in the conversation. The bot offers the closest package and notes that a custom retainer can be discussed on a discovery call. It never invents a price for a non-existent package, and the agency keeps control of bespoke quotes.

Yes. The bot can pass the matched assistant slug and package ID as query parameters to your intake form or Stripe checkout, so when the prospect arrives the relevant fields are prefilled. If you use Gravity Forms or WPForms, SleekAI's tool calls can also submit the form directly with the captured details when configured.

Yes. SleekAI Multibot lets you run separate bots scoped by URL or page template. A bot on the bookkeeping landing page can be instructed to surface only finance VAs and quote only the bookkeeping packages, while the executive support page bot focuses on EA work. Each bot has its own instructions and presets.

Conversations are logged with model, tokens, and origin page in the SleekAI conversation table, but you control retention. You can configure log retention to clear after a set number of days, exclude logged-in user IDs, or strip specific fields. Visitors can also be given a clear opt-in notice in the initial bot message.

Yes. Even with a single assistant, the scoping benefit still applies. The bot can speak to your service packages, quote rates accurately, and explain what's included so prospects qualify themselves before booking a discovery call. It also handles after-hours leads when you can't reply in real time.

GPT-4o-mini or Claude Haiku handle matching and quoting at a low per-conversation cost, usually under one cent. For complex multilingual briefs, GPT-4o or Claude Sonnet read context more carefully. You can configure different models per bot so the top-of-funnel bot uses Haiku while the sales bot uses Sonnet.