How an answer comes together
My-Chatify uses a method called RAG. For every question, the bot first looks up the relevant passages in your documents. Only then does the AI model turn them into an answer. The knowledge comes from your sources; the model just makes the sentences readable.
Each assistant has its own separate database for this. What you give one assistant never shows up in another, and it is not used to train AI models. The answer appears word by word as it is written, so the customer can see straight away that something is happening.
If the bot finds nothing suitable in your sources, it says so politely and points the customer to your support team, for example. There is no absolute guarantee with AI. That's why a thorough test before launch is part of the job.
- Searches your documents before every answer
- Its own isolated database per assistant
- Answers appear as they are written
- Says openly when it doesn't know



