Every AI chatbot answers questions. That's the low bar. If we hadn't done anything more than wrap that in a school-shaped interface, we would have built a slightly nicer chat window. So this note is about the four decisions that separate the eZine mentor from a chat window — and why they matter enough that we're describing them on a marketing page.
1. It reads your Business Progress Record before it answers.
A generic chatbot treats every question as blank-slate. Ask about pricing and you get the general pricing lecture — good in the abstract, useless in your business. The eZine mentor starts every response from your own Business Progress Record: your stage, your industry, your last month's data, the pricing rule you already wrote down, the artefacts you've saved. The generic answer becomes the answer to your question, in your business, with what you already have. If you ask "should I raise my price?" the response references your own margin history — not a hypothetical one.
2. It cites the lesson behind every claim.
A chat window can be confidently wrong. The mentor is engineered to cite the specific lesson, guide or reference behind each substantive claim. If the source isn't there, the mentor says so and doesn't invent one. Made-up citations are the failure mode of AI-in-education this decade; we wanted the citation to be a hyperlink you can click, not a name that sounds convincing.
3. It escalates regulated questions.
Legal, tax, licensing, health, regulated-financial. Every one of these has a domain of qualified humans, an insurance requirement, and stakes small enough to break a business if the answer is wrong. The mentor is trained to recognise those questions and route them to human review — a qualified professional in your jurisdiction, or an eZine staff member who can hand off appropriately. It doesn't try to answer them. It says who to ask, and why. That is a deliberate feature, not a limitation.
4. You can correct it — and the correction feeds forward.
Every mentor response carries a "flag for correction" control. When you use it, an eZine reviewer reads the exchange, and where the correction sticks, the underlying retrieval is updated so the next learner doesn't run into the same wrong answer. This is the loop that turns a good static model into a school that gets better. A chat window does not have this loop. We have chosen to make it visible.
Why write this note at all?
Because we're launching in a market where AI features are described with adjectives — smart, personalised, cutting-edge — and rarely with mechanisms. We wanted the mechanism written down. If a competitor's AI does these four things too, we won't be the differentiator. But the marketing page will still read the same, and this note will still be true.
— eZine Trades & Business School · Nigeria launch, 2026