Frequently asked questions

Straight answers, including about what isn't built yet.

What is AnswerBnB?

An AI Operations Platform that connects a business's own knowledge, its live operational context (reservations, orders, tickets), real conversations, and a controlled layer of automation and human escalation — built to work across property, ecommerce, customer support, and ISP/NOC operations on one shared architecture.

Is AnswerBnB a chatbot?

No. A chatbot answers questions. AnswerBnB resolves who and what a conversation is about first, retrieves only knowledge scoped to that context, and chooses between answering, clarifying, escalating, or flagging a conflict — a chat interface is only the visible surface of that.

How does AnswerBnB know which property, customer, or order a conversation is about?

Every conversation is tied to a real organization and, where applicable, a specific property, reservation, order, or ticket at the data level — this is enforced in the application and database, not inferred loosely from message text alone.

How does the Knowledge Engine work?

Each property or service can have multiple knowledge sources — manual entries and crawled URLs. Sources are fetched respecting robots.txt, extracted into text, deduplicated by content hash, and stored as chunks the AI can retrieve from — scoped to the property that owns them.

What happens when the AI doesn't know the answer?

It says so and escalates to a human with the full conversation attached, instead of guessing. Confidently answering the wrong thing is treated as a worse outcome than an honest "I don't know."

What happens when information conflicts?

Detecting conflicting knowledge (for example, one source saying check-in is 3 PM and another saying 4 PM) and surfacing it rather than silently picking one is part of the Knowledge Engine's designed architecture, being built out on top of the current single-source-per-fact foundation.

Can a human take over from the AI?

Yes — escalation to a human is a core part of the design, with full conversation and grounding context carried over, not a cold handoff.

Can AnswerBnB connect to systems we already use?

That's the intent of the architecture: a normalized adapter layer designed to sit around property-management, ecommerce, CRM, help-desk, and ISP/NOC systems rather than requiring you to replace them. See Integrations for exactly what's built today versus designed for the future.

What property-management platforms are relevant?

Systems like Guesty, Hostaway, Hospitable, Lodgify, OwnerRez, and Smoobu represent the kind of PMS ecosystem AnswerBnB's architecture is designed to connect with.

What ecommerce platforms are relevant?

Platforms like Shopify, WooCommerce, and BigCommerce represent the kind of ecommerce ecosystem the architecture is designed to connect with.

What CRM and help-desk systems are relevant?

Systems like Salesforce, HubSpot, Zendesk, and Intercom represent the kind of CRM/help-desk ecosystem AnswerBnB is designed to sit around.

What ISP and NOC systems are relevant?

Systems like Splynx, Sonar, UISP, and MikroTik-based networks represent the kind of ISP/NOC ecosystem the architecture is designed to connect with.

Are all the platforms listed on the Integrations page currently integrated?

No. Every vendor name is labeled honestly — Available, In Development, Architecture Compatible, or Planned — and as of today, no third-party vendor connector is live. Naming a platform describes the ecosystem we're building toward, not a claim that a live connector exists.

How does knowledge stay fresh?

Sources track when they were last crawled and last verified, and content is hashed so an unchanged re-crawl doesn't create duplicate knowledge. A fuller freshness/staleness status model is part of the Knowledge Engine's ongoing build-out.

Can humans correct the AI?

Yes, and doing so is designed to be able to feed back into the organization's approved knowledge — so the same gap doesn't need correcting twice — rather than only affecting the single conversation.

Can AnswerBnB automate workflows?

The platform's automation is built around a strict principle: it may only execute actions it has been explicitly authorized to take. A general-purpose automation/journey engine (trigger → condition → AI decision → action) is architecture and roadmap today, not a shipped feature.