Glossary
Inference
Inference is the moment a model runs to produce a reply or a tool request. Cost, latency, and the token ceiling all happen here.
Also called استنباط مدل
Training a model is rare and expensive; inference is what repeats with every sales-manager question. That is why a good system keeps repeated work and exact calculation off the model.
Which model to use for inference is an operational choice: a stronger model is not automatically better at counting invoices.
Let’s make your processes agentic
Where do we start?