Exode team
11:40 pm. A student messages you: "where's the second assignment?" In the morning, four more with similar questions. At lunch someone asks about something covered in lesson one.
You answer. Because if you don't, the student won't do the assignment — and soon won't log in at all.
The problem is that it doesn't scale. Thirty students is an evening of messaging. A hundred and you're no longer teaching.
What a course knowledge base is A course knowledge base is where your materials and answers to student questions live: the program, deadlines, access terms, instructions, and breakdowns of the questions that come up on each topic. Its job is to answer a student at the moment they get stuck, without waiting for a teacher or curator. On the Exode platform an AI assistant sits on top of the knowledge base: the student asks in plain words and gets an answer assembled only from the school's own materials.
Student questions are your school's ceiling
The logic is simple: you can take on as many students as you can answer.
The usual ways out:
- hire a curator — eats into margin, especially on cheaper courses;
- stop answering — students drop off and completion falls;
- don't open the next cohort — don't earn.
There's a fourth: make most of the questions get answered without you.
How it works in Exode
The student asks a question inside the platform and gets an answer in seconds.
- The answer is assembled only from your files. Not from the internet, not from the model's general knowledge. What you uploaded is what it answers from.
- You don't spend time on repeat questions. It frees up your school's most expensive resource — your time.
- Nobody interrupts you at night or on weekends. The student gets an answer while they're actually studying.
- The student doesn't wait. That matters more than it sounds: waiting is what usually ends the study session.
Why this affects more than your comfort
A student gets stuck late at night. They get an answer — they finish the assignment. They wait until tomorrow — they postpone. A week later they don't come back at all.
Students who finish leave reviews, become case studies and buy the next course. The ones who drop out do none of that. So response speed isn't customer service — it's next cohort's revenue.
What to put in the knowledge base
Start with what people already ask:
- the course program, dates and deadlines;
- access, payment and instalment terms;
- instructions: how to submit an assignment, where to find the recording, what to do if a video won't play;
- breakdowns of common mistakes by topic;
- handbooks, notes and extra materials.
What AI won't do
Stated plainly, so expectations stay realistic.
It won't assess creative work — that needs your eye. It won't replace a personal message to a student who's stuck and discouraged. And it won't fix a badly explained topic: if a lesson is unclear, questions about it will keep coming.
The right framing: AI closes the repeat questions, you stay where a human is needed.
Where to start
- Collect the questions students asked during your last cohort.
- Pick the 15–20 most frequent ones — that's the content of your knowledge base.
- Upload the course materials and the answers to those questions.
- On the next cohort, look at what still gets asked: those are the topics to rewrite inside the course itself.
That last step is usually underrated. The list of frequent questions is a ready-made plan for which lessons to improve before your next launch.
To see how it works, you can try our platform free for 7 days. Get in touch — https://t.me/exode_biz_support