Your data, usable by AI. RAG in WEBCON

LLMs have made finding information fast. They speed up the work that follows, too. But a model only knows what it learned from public data. So what if you could pair that speed with the knowledge your organization has already built? That's what RAG is for. 

What is RAG?

RAG (Retrieval-Augmented Generation) connects LLMs to external data sources. That lets them pull in relevant information instead of relying only on training data. Inside an organization, RAG can work as a retrieval layer that connects AI to your business data.

How does RAG work in WEBCON?

When configuring an AI agent, the developer picks one or more files the agent can use. These come from dictionary processes, the place where your organization stores standing reference material like policies, procedures, and guidelines. The business owns and maintains those files, so the knowledge stays current without a separate technical project.

When an AI agent asks a question, the system doesn't just match keywords. It performs a semantic search across your data, meaning it understands that "vacation" and "time off" and "absence" are related concepts, even if they use different words. The system gathers the most relevant pieces of information, bundles them together, and passes that context to the AI model.

Who can access the data?

Like in any other cases when it comes to WEBCON, the access control applies here too. Administrators configure which files RAG can search, so if certain content should stay out of reach for specific roles, that boundary is respected. RAG never surfaces content someone shouldn't be able to see.

What your teams gain

• Faster answers. An agent acting for them, asks a question and gets the relevant procedure back, without opening a handbook or searching for a shared drive.
• The same answer for everyone. A new hire, an experienced employee, and an AI agent all draw from one source, so nobody works off an outdated or secondhand copy.
• Solid fundaments for AI use. Your agents can run on knowledge the business already owns and maintains, so what they return stays as current as your policies and procedures.

Easy access to your data is how you win

Every company knows embedding AI into your business processes is a way to go in 2026 and beyond. But it has to have access to your company’s business context and data, if you want it to be effective.  

GTX Service did the groundwork first. Results? 48 times faster time in the customer service center and 110 times lower cost per CSC request.