Put Your Knowledge Base Where People Already Are

There’s a gap between having a knowledge base and people actually using it. You can write the best articles in the world, but if someone has to stop what they’re doing, open a new tab, log into a separate tool, and search from there, a lot of them won’t bother. They’ll ask a colleague. Or guess.

This is especially true for two groups: customers who need help but don’t want to dig through a support portal, and internal teams who need quick answers while they’re in the middle of something else. A CRM. A ticketing system. A call.

The fix isn’t better documentation. It’s putting a knowledge base chatbot where people already are.

A widget is a small piece of code you add to any website or internal tool. It gives people access to your knowledge base without leaving the page they’re on.

There are two ways it can show up. As a floating AI chatbot in the corner of the page, the kind you’ve probably seen on support sites. Or as an embedded panel built right into a page, like a dedicated help section on your website or a knowledge panel inside your CRM.

Either way, the person gets search, AI chat, or both, right where they’re already working. No new tab. No separate login. No context switching.

Three modes, your choice

Not every situation needs the same thing. Sometimes people want to browse and search. Sometimes they want to ask a question in plain language and get an answer from an AI chatbot. Sometimes they want both.

That’s why we built three widget modes that you choose when setting up each widget.

Search only. The widget shows a search bar. People type a query, get ranked article results, and can click through to read the full article. No AI involved, no AI costs. This is good for teams that want fast, direct access to specific articles without the overhead of an AI chatbot response.

AI Chat only. The widget shows an AI chatbot interface. People ask questions in their own words and get a conversational answer drawn entirely from your knowledge base. The chatbot finds the relevant articles, pulls the information together, and gives a clear, grounded response. It doesn’t use the internet or make things up. If the answer isn’t in your content, it says so. Good for customer-facing use where you want people to get answers without having to know which article to look for.

Both. This is the default. The widget gives people a toggle between search and AI chat, so they can switch depending on what they need. Want to browse? Search. Want to ask a specific question? Ask the chatbot. It covers both use cases without you having to choose one over the other.

Citations: on or off

When the AI chatbot gives an answer, you can choose whether it shows source links or not.

With citations on, every answer includes clickable links back to the articles and documents the information came from. The person can verify the answer, read the full article, or dig deeper. This builds trust, especially for internal teams where accuracy matters and people want to check the source.

With citations off, the chatbot gives the same quality answer but weaves it naturally, without referencing specific articles. This creates a smoother, more conversational experience. Good for customer-facing widgets where you want clean, helpful answers without exposing your internal article structure.

It’s a toggle in the settings. You can change it any time.

External vs internal

There’s an important distinction in who sees what.

External widgets are for your customers or the public. You add them to your website, your help page, or your product. They only show articles you’ve marked as public. Your internal processes, policies, and team-only content stay hidden. Someone browsing your website will only see what you’ve chosen to make visible.

Internal widgets are for your team. You add them inside your CRM, intranet, internal portal, or any tool your staff use daily. They show everything, both internal and public articles. They’re domain-locked, so they only work on the websites you’ve approved. And they’re email-gated: the first time someone uses the widget, they enter their work email to confirm they’re a real user in your system. After that, it remembers them.

The difference matters because it means you can run both from the same knowledge base. Write an article once. Make it public, and it’s available through your customer-facing chatbot widget. Keep it internal, and it’s only visible to your team through the internal widget. Same content, different doors.

It picks up your branding automatically

The widget pulls your company colours and logo from your KnowledgeScout settings. So when it appears on your website or in your internal tool, it looks like part of your product, not a third-party add-on. No design work needed on your end.

Where this actually gets used

The most obvious use case is an AI chat widget on your website. Someone has a question, they click the chat bubble, they get an answer from your knowledge base. That’s useful and it reduces support tickets.

But the more interesting use case is internal. Picture a contact centre agent on a call. They need to look something up. Instead of leaving their CRM to search the knowledge base in another tab, the chatbot widget is right there inside the CRM. They type the question, get the answer, and keep going. No tab switching. No lost time.

Or an operations team using an internal portal. Someone needs to check a process. The widget is embedded in the portal page. They search, find the article, and they’re done.

The pattern is the same every time. Bring the knowledge to where the work happens, instead of making people go find it.

The honest bit

A widget is only as good as the content behind it. That’s why ours sits on top of a full knowledge management platform, not just a content store.

Editors can schedule review dates so articles get checked before they go stale. Readers can leave comments, vote on whether content was helpful, and flag articles that look outdated or wrong. Those signals go straight back to the people responsible for maintaining the content.

The result is a feedback loop. The widget surfaces your knowledge to the people who need it. Those people tell you when something is off. Your editors fix it. The next person who asks the same question gets the right answer.

Most AI chat widgets are just a delivery layer on top of static content. Ours is connected to the tools that keep that content alive. That’s the difference between a knowledge base chatbot that works on launch day and one that’s still giving accurate answers six months later.

The KnowledgeScout Team