AI agents & agentic AI
Meet your AI agent
A chatbot answers from memory. An agent takes steps — it opens your own Studio analytics, reads what's there, then answers how your courses are doing across every school.
What is an AI agent, in plain words?
No jargon. Two short answers and one real example.
A chatbot talks. An agent takes steps.
A chatbot answers from what it already knows. An agent works like a colleague with your filing cabinet open: it opens the right analytics page, reads the real rows, then answers.
“Agentic AI” is AI that acts within limits you set.
Agentic means the AI can act — run a lookup, go back for a detail — within limits you set. In EduGears AI Studio, the agent can read your own courses, and reading is all it can do.
Example
What that looks like
How is my Circuits course doing across schools?
Circuits & Signals is live in 7 institutions, with 412 learners across 9 cohorts. Completion is 63% overall, but two cohorts sit under 30% — both redeemed the code in the last three weeks. Learners stall most often in Module 4, “Transient response”. Want the learner table for those two cohorts?
Where this came from
Course Analytics · 9 cohortsCog opened the same Course Analytics page you can open yourself — click the source to land on it, filters applied.
In your workspace
How the assistant works in EduGears AI Studio
Cog is the sparkle in the corner of every page. Ask in plain words; it answers from the analytics you already have.
Ask across every school at once
One question covers every institution that redeemed your code — no reports, filters or spreadsheets.
Find where learners get stuck
“Which module loses people?” “Who never started?” Cog answers from your cohort and learner tables.
Follow a code's reach
Which institutions run a course, when they redeemed it and how far learners have got.
Answers you can open
Every figure links to the analytics page it came from, same filters applied.
Guardrails, not promises
The limits are built into how the agent runs — not merely written into its instructions.
Read-only, always
Cog can look, never change: no course edits, no code changes, no records touched, no messages sent.
Your own catalog, nothing wider
Only courses your organization owns. An institution's private data stays theirs; their instructor sees only their section.
Every number is traceable
Answers name their source page and use the same analytics your team reads. Missing data? Cog says so.
Your switch, in Settings
Admins turn course-data questions on or off for everyone in Settings; Cog keeps answering everything else.
Questions publishers ask
What is an AI agent?
An AI that takes steps to answer you: it looks things up — here, in your own analytics — and answers from what it found, not from memory.
Can it change my courses or my access codes?
No. Cog is read-only by construction: it has no tool that edits, publishes, regenerates a code or deletes anything.
Can it see an institution's private data?
No. Cog answers only about your organization's courses and the learners taking them. Everything else in a school's LMS stays with the school.
Where do the numbers come from?
From the same Studio analytics your team reads — which is why its figures match the page, and every answer links back to it.
Do we have to set anything up?
No. Cog is built in — click the sparkle and ask. Admins can switch course-data questions off. Messages cost 2 AI credits each (very long messages cost 2 more for every extra 10,000 tokens), or run on your own AI keys.
Ask about your own courses
Cog is already in EduGears AI Studio. Ask how a course is doing across the schools that run it — or start free and publish your first one.