Pinly uses AI to help your team move faster. We think carefully about what that means for the information you trust us with, and we would rather explain our approach plainly than ask you to take it on faith.
There is a quiet assumption in a lot of modern software that giving a model your data is the same as giving it to the company that built the software. We do not see it that way. Your course's information is yours. We use AI to do specific jobs, and the question we ask ourselves first is not "will the model do well with this", but "does this information need to be here at all?"
For the jobs AI does well, it does not need to know who your team is or what your course is called. It needs to know the shape of the work, when it happens, and what conditions surround it. Pinly handles the translation. You see your team by name on every screen and in every reply, while the model sees the operational picture without the labels that identify any specific person or place.
There is a meaningful difference between the public ChatGPT, Gemini, or Claude apps and the enterprise versions of those services. The free apps may use your conversations to train future models. The enterprise services are contractual, and do not. Pinly uses Google Vertex AI in the European Union under those enterprise terms, which means your prompts are not used to train Google's models and not visible to any other customer. Google keeps prompts only briefly, to run its service and to detect abuse of it.
We would rather talk through it on a short call than leave it sitting unresolved.