How findIQ uses AI

The two types of AI in findIQ, the findIQ diagnostic AI and large language models, what each one does and where to find details.

Written By Tommy Giesbrecht

Last updated About 3 hours ago

Two types of AI

findIQ combines two approaches that do different jobs:

	findIQ diagnostic AI	Large language models (LLMs)
What it does	Guides troubleshooting from symptom to cause and learns from real service cases	Quinn: turns documents, expert knowledge and videos into findIQ content
Type of model	Model-based, self-learning statistical methods	Generative language models
Developed by	findIQ	OpenAI (via Microsoft Azure)
Where it runs	findIQ servers in the EU	Microsoft Azure (Europe); data is stored on findIQ servers in the EU
Human control	Changes to the knowledge base are suggested and need approval by an Editor or Administrator	Acts with the permissions of the signed-in user, deleting is blocked
Availability	Core of findIQ	Enabled per organisation

Why two approaches?

Troubleshooting on machines needs reliable, traceable answers. That is why the diagnosis itself does not rely on an LLM: the findIQ diagnostic AI works only on the knowledge your experts recorded and does not invent causes.

LLMs are used where language matters: capturing knowledge from documents and conversations, and operating findIQ through a chat. Even then they work on the structured findIQ data model, not on a loose file store, so answers can be traced back to their source.

Using your own AI system

You can also connect your own AI system, for example Microsoft Copilot, Claude or ChatGPT, via the findIQ MCP server. Provider and model are then your choice. See Using findIQ in Microsoft Copilot and other AI systems.

Was this helpful?