Frequently asked questions
A generic chatbot retrieves a chunk of text and hopes it was the right one. Mention does something different: it builds a curated, dependency-ordered learning path with knowledge checks and per-member progress tracking. Answers are constrained to your team's vocabulary. When sources conflict, you decide which is correct; when an answer is wrong, you correct it inline and the change propagates to everyone.
No. You curate, the AI writes. You describe an audience in a chat about what they do and what they need to learn, and the AI proposes processes and concepts pulled from your existing docs, meetings, and chats. You accept, rename, or skip. Articles are generated on demand from the latest source material, so they don't go stale the moment you ship them.
You correct it inline, and the correction propagates to every learner immediately. The old answer is gone. When sources disagree, Mention surfaces the conflict and asks you to decide rather than picking arbitrarily.
We don't train models on your content. We store structured analyses tied to concepts (what each source says about a given idea), and we do not retain your raw documents.
One focused working session for the first audience: connect a source, run through the curation conversation, invite your team. Learners can start working through processes the same day, at their own pace. Adding more audiences later is faster because sources are already connected and the glossary builds on what's already there.
Sources refresh on a schedule, so updated docs flow into the next article generation automatically. Reopen the curation chat any time to add, rename, or remove processes, and the playbook reflects the change immediately. Switching tools means disconnecting the old source and connecting the new one; your audiences, playbooks, and glossaries aren't tied to a specific integration.