Continuous Training

This feature improves your assistant's responses after it is already online, based on real questions from your customers.

How it works

The system learns from daily use. Imagine a customer asks for the opening hours of a new office you didn't include: the assistant won't know how to answer, but the system will detect this gap.

Before go-live, we cover all critical points through initial training. Once online, real interactions show what is missing or what needs clarification. You don't need fixed check-up routines: improvement happens naturally as customers use the service.

What you need to do

You don't have to follow mandatory maintenance programs. Only intervene in two cases: 1. When you notice the assistant gave an imprecise answer. 2. When company information changes, such as a product price or an internal procedure. In this case, follow the guide on su casi.

What glacom® does

The glacom.AI team monitors interactions, detects gaps in responses, and tunes the system to make the assistant increasingly accurate.

Limits

The assistant cannot invent information that does not exist in your sources. If a customer asks for data that was never provided to glacom®, the system cannot answer correctly until the information is added.

Frequent problems

• Do I need to dedicate time every week to training? → No, the system refines itself through use and glacom® intervention.
• What happens if I change a price tomorrow? → You must report the change to update the knowledge base.
• How do I know if the assistant is making a mistake? → You can check the interactions or wait for feedback from glacom®.

From the app

Not available from the app: knowledge management and response analysis take place on the web platform.