"Deploy in minutes" usually means a demo bot answering generic questions, not a production agent your customers can rely on. Real deployment is training, testing, and getting the handover right, and it takes a bit longer than a marketing page suggests, for good reason.

What actually happens between signing up and going live

Deployment isn't a single switch. It moves through a sequence: account setup, knowledge configuration, agent configuration, testing, and only then, deployment.

Account setup. After an initial demo and onboarding conversation with the business, the account is created and the business receives an email invitation to access the platform and set a password. This part is quick and largely administrative.

Creating the agent and giving it knowledge. The business names the agent and gives it the material it needs to actually help customers: uploaded documents, or a connected website, or both. This is the step that determines everything downstream. An agent is only as good as what it's been given to work with, and a business can keep adding source material over time as they notice gaps in what customers ask about.

Configuring how it behaves. Once the agent has knowledge, the business configures its appearance, tone, and behaviour from the dashboard, deciding how and when it should interact with customers and how it should represent the business. This is where a generic agent becomes one that actually sounds like the business it's serving.

Testing, properly, before customers see it. This is the stage that shouldn't be rushed, and it's also the stage most "minutes to deploy" claims skip past. Before anything goes live, the agent should be run through chat, voice, and search playgrounds, testing real questions, finding the gaps in its knowledge, and refining its behaviour, so the first real customer conversation isn't also the first time anyone's checked whether the agent gets it right.

What can happen on day one, and what shouldn't be rushed

Account setup, the initial knowledge upload, and basic agent configuration can genuinely happen on day one. There's no reason to stretch that part out.

Knowledge validation and testing are different. Uploading a document and switching an agent on doesn't guarantee it'll give accurate or useful answers to the questions your actual customers ask, worded the way they actually ask them. Skipping this step is exactly how a fast-looking deployment turns into a slow, expensive one later, once customers start hitting the gaps a proper test would have caught.

A realistic timeline, not an arbitrary one

For a straightforward business with well-organised information already available, initial setup and testing can realistically be completed within one to two business days. A business with a larger knowledge base, more integrations, or more specialised workflows will reasonably take longer, and that's not a failure of the process, it's a reflection of genuinely more to configure and test properly.

The point isn't to promise the fastest number on the page. It's to be honest that "live" and "ready" aren't always the same day, and the gap between them is exactly the work that determines whether the agent is actually useful once real customers start talking to it.

If a vendor promises "live in minutes"

That's not necessarily untrue, but it's worth asking exactly what "live" refers to. A few questions separate a genuine timeline from a marketing one:

• Does "live" mean the agent answering demo questions, or answering real questions from your actual knowledge base?

• Has the agent been tested against your material before real customers reach it, or is testing something you do live, in production?

• What happens in the first week, does someone review what the agent got wrong and adjust it, or is it left to run as configured on day one?

• Is there a difference in the timeline between a straightforward business and one with a larger or more specialised knowledge base, or is every business quoted the same number regardless?

A vendor who answers these plainly, rather than repeating the headline figure, is usually one whose "minutes" claim will hold up once you're the one relying on it.

The distinction that actually matters

The right question isn't how quickly an agent can be switched on. It's how quickly it can be trusted with a real customer conversation. A fast deployment built on thin knowledge and no testing can create more problems than it solves: wrong answers, missed handovers, and a customer experience that undoes whatever time was saved getting the thing live. A deployment that takes a couple of extra days but is properly tested is the faster option in every way that actually matters, because it doesn't need to be redone once customers start finding the gaps.