AI Tutor — the full guide

1. What the AI Tutor is

The AI Tutor is the part of CampusTutor that explains your coursework one-on-one — like a private lecturer who has read your notes, your institution's past questions, and your current level, and is available at 2am the night before a test.

It is not a generic chatbot. Every response is shaped by three things: the course you're in, the topicyou're studying, and your academic profile(university 100L–500L, polytechnic ND1–HND2, or College of Education NCE1–NCE3). A "explain X" question in 200L gets a different answer than the same question in HND2 — same accuracy, different depth, different language.

2. Plans & message limits

The AI Tutor is included in every plan, but daily message limits differ. One messageis one question you send (the AI's reply is free — you only spend when you ask).

When you hit the Free-plan daily limit, you'll see "That's your 10 free messages for today" and the send button locks until midnight (your local time). Voice input and image uploads still work on Plus and above.

3. Starting your first session

1. From the dashboard sidebar, click Tutor Mode to open the course picker.
2. If you don't see any courses yet, click Add Course in the top bar. Search by course title or code (e.g. "MTH 201" or "Engineering Mathematics") and pick your institution from the dropdown so past questions match your school.
3. Click any course tile to enter the course workspace.
4. Pick a topic from the left panel — or click Ask anything about this course to start a free-form chat that spans all topics.
5. Type your question and press Enter. The AI streams its answer back live.

4. The Tutor Mode workspace

Tutor Mode has three regions:

• Left — Topic Panel. Every topic in the course, sorted by recommended study order. Each topic shows a mastery bar (0–100) and a status dot (not started / learning / understood / mastered). Hover any topic and you'll see a Brain icon — click it to launch a Topic Quiz directly.
• Centre — Tutor Chat. The conversation. Streaming responses, math rendering, image previews, and copy/listen buttons on every reply. Scroll up at any time to revisit earlier explanations — your history is preserved per topic.
• Bottom — Learning Controls. The input area, plus four colour-coded action chips (Say it simply, Show examples, Break it down, Check my understanding), a paperclip for image upload, and a mic for voice input.

On mobile, the Topic Panel collapses into a drawer (tap the menu icon). The chat takes the full screen.

5. The four explanation modes

Above the input box you'll see four pill buttons. They're shortcuts that send a pre-tuned instruction to the AI so you don't have to type "explain again, simpler" yourself. Tap one and the next response is rewritten in that style.

Say it simply

Strips out jargon and translates the concept into plain everyday language. Use this when an explanation feels heavy or when the textbook definition doesn't click yet.

Show examples

Pulls in three to five concrete, real-world examples. The AI prioritises scenarios you'd actually meet in your field — engineering students get machine examples, medical students get patient examples, accounting students get journal-entry examples. Best for memorisation and for spotting how a concept shows up in exam questions.

Break it down

Walks through the concept step by step, explaining the reasoning behind each step. The default for mathematical, computational, or process-based topics (algorithms, integration, balancing equations, drug calculations).

Check my understanding

Flips the dynamic — the AI asks you a question and waits for your answer before continuing. It then gives kind, specific feedback. This is the closest thing to having a tutor quiz you and is the fastest way to find out whether you actually understand or just feel like you do.

6. Topic sessions

When you click a topic from the panel, you enter a structured topic session — different from free-form course chat. The AI runs a four-part arc:

1. Activate. A short opening question to surface what you already know about this topic. Your answer (or skip) tells the AI where to start.
2. Explain. The core teaching, delivered in chunks — each chunk a digestible idea you can stop and re-read.
3. Check. After roughly every fourth chunk, an Understanding Check pops up — a multiple-choice question grounded in what was just explained.
4. Consolidate. A wrap-up that ties everything together and asks you to self-assess how confident you feel.

You can leave a topic session at any time — your progress, chunks, and check results are saved. Coming back later resumes where you left off.

7. Understanding checks

Understanding Checks are short multiple-choice questions the AI inserts between chunks of explanation. They are not graded against your GPA — they are learning moments. Their job is to catch confusion early, while the explanation is still fresh.

When you pick an answer:

• If correct, you get one of eight rotating encouragement messages and the session continues.
• If incorrect, the card shows the right answer in a calm slate colour (not red or amber — we don't want it to feel like punishment) along with a one-line explanation of why.
• Your check-in correct/total ratio is fed back to the AI as session memory, so later explanations adjust for the things you keep missing.

8. Uploading notes & past questions

CampusTutor accepts PDF, DOCX, TXT, and image files (JPG, PNG, WEBP, TIFF, BMP) up to 20MB per file. Uploads happen from the course workspace — look for the Upload button in the top toolbar.

What to upload:

• Lecture slides & notes — the AI cites them directly and uses your lecturer's phrasing, not generic textbook prose.
• Past questions (your school's). The AI surfaces these when you practise and uses them to ground explanations in the way your school actually tests.
• Recommended textbooks — chapters or extracts. Don't worry about scanned, blurry pages; OCR is built-in.
• Your own handwritten notes (scan or photograph them). OCR handles handwriting up to about 80% legibility.

9. Snapping a question with your camera

The paperclip button next to the input lets you attach an image (or take one with your camera on mobile). The AI reads the image and answers grounded in what it sees — perfect for:

• Past question papers you photographed during a study group.
• A worked example from the textbook you don't understand.
• A diagram, schematic, or chart you need explained.
• A handwritten note where you're not sure if you copied the formula correctly.

Image size cap is 5MB per attachment. If your photo is too large, crop or compress before attaching. You can send text plus an image in the same message — the AI uses both.

10. Voice input & Listen mode

The mic button to the left of the input transcribes speech into text. Tap once to start, tap again to stop. The transcript appears in the input box — review it, edit if needed, then send. Voice input requires Plus or above.

Every AI reply also has a Listenbutton. Tap it once and your device's built-in voice reads the answer aloud (good for revising on the road). Long-press it (~600ms) and a smoother neural voice takes over for a more natural reading experience.

11. Quantitative step-by-step

When the AI detects that your question is computational (algebra, statistics, calculus, accounting calculations, physics, engineering problems, financial maths), it automatically generates a structured step-by-step solution alongside the chat answer.

The step-by-step panel appears under the AI's reply with each step numbered, each formula rendered properly with LaTeX, and the final answer boxed. You can collapse the panel if you only want to see prose, or copy any individual step to your notes.

12. Area of Concentration (AOC)

Each course in your workspace has an Area of Concentration setting — this is where you tell the AI which topics your lecturer has flagged as high-importance for the upcoming exam. Open any course, find the AOC section in the course header, and add the topics your lecturer emphasised.

What AOC does once set:

• The AI biases its examples and depth toward your AOC topics.
• Course Tests and Exam Simulations can be set to bias questions toward AOC topics (a toggle appears when starting the assessment).
• The Readiness Score (see the GPA & Academic Tools guide) weights AOC mastery more heavily.

13. How mastery is calculated

Each topic has a Mastery Scorefrom 0 to 100, shown as a small bar next to the topic name. The score updates automatically as you study — you don't need to refresh anything.

It moves based on:

• Understanding Check results in topic sessions (biggest single contributor).
• Topic Quiz, Course Test, and Exam Simulation results — graded by an adaptive engine that weights each question by difficulty.
• Self-assessment at the end of consolidate phases (a smaller contributor).
• Time decay — if you haven't touched a topic in a few weeks, mastery slowly drifts down, prompting a refresher.

Mastery status thresholds:

• Not started (0) — grey dot.
• Learning (1–59) — amber dot.
• Understood (60–84) — sky-blue dot.
• Mastered (85+) — green dot.

14. Postgraduate features

If your academic profile is set to Postgraduate, the AI Tutor unlocks two extra modes for research-stage students:

• Proposal phase. The AI helps you sharpen research questions, scope objectives, critique your methodology, and stress-test your conceptual framework.
• Viva phase (mock viva). Toggle viva practice mode and the AI switches into examiner mode. It asks the kind of probing questions a real panel would — about your contribution to knowledge, your methodological choices, and the limits of your findings — and gives you a chance to defend each one.

Postgrad students should also fill in their research area in the academic profile — the AI uses it to make examples and viva questions domain-relevant.

15. How to get maximum results

1. Upload before you ask. The single biggest quality jump comes from uploading your lecturer's notes. The AI uses your school's wording, your lecturer's preferred examples, and your institution's past questions.
2. Use the four mode chips, not custom prompts. "Explain again, simpler" works, but tapping Say it simply is a tuned instruction that's been A/B tested to give better results.
3. Always let it check you. After any topic explanation, tap Check my understanding. If you get the answer right, you knew it. If you got it wrong, now you know which detail tripped you up — before the exam, not during it.
4. Don't skip the Topic Panel. Free-form course chat is fine for quick questions. But for actual learning, click into a specific topic — the structured Activate → Explain → Check → Consolidate arc is what builds mastery scores and feeds GPA prediction.
5. Set your AOC weekly. Spend 30 seconds after each lecture updating your Area of Concentration. By exam week, your AOC is the high-yield study plan.
6. Use voice on the bus. Ask questions hands-free, listen to replies. Even 15 minutes between lectures adds up.
7. Snap past questions immediately. When a senior shows you a past question, photograph it and upload to the relevant course. Future you will be grateful.

16. FAQ

Does the AI Tutor work offline?

Partially. Previously-loaded responses remain readable offline (handy on slow campus WiFi), but new questions need a connection because they stream live from the AI. The PWA caches your last session so you can re-read it without data.

Why does the AI sometimes say it doesn't know something?

Because honesty matters more than hallucination. If your question is outside the course material the AI has, it will tell you and suggest uploading something more relevant. Better than a confident wrong answer.

Can I export a conversation?

Each AI reply has a Copy button — tap it to copy the answer to your clipboard. A full conversation export to PDF is on the roadmap.

Will the AI write my essay for me?

The AI Tutor is built for learning, not ghost-writing. If you ask it to "write my essay" it will offer to teach you the structure, give you a skeleton you fill in, or coach you through your own draft. Essay grading (which marks your essay and tells you what to improve) is a separate feature available on Plus and above.

What if my question is in pidgin or Yoruba/Igbo/Hausa code-switching?

The AI handles it. You can mix English with Naija pidgin or any of the three major languages and it will respond in English (or in your style if you ask). It will also explain technical terms in your local language if you ask it to.

How is this different from ChatGPT?

Three things: (1) it's grounded in yourcourse material and your institution's past questions, so answers are tailored to how your lecturers actually test; (2) it tracks your mastery per topic and adapts to where you're weak; (3) it's built for the Nigerian tertiary system — it knows what 100L vs HND2 vs NCE3 means and adjusts accordingly.