A Step-by-Step Guide for Instructors Using Canvas LMS
Tutorify AI · Instructor Onboarding Guide · 2026
1 / 35
?
Overview
What is Tutorify?
Tutorify is an AI-powered educational platform that integrates directly into your
Canvas LMS. It helps instructors manage courses, engage students with an intelligent chatbot,
generate quizzes, and automatically grade assignments — all from within your Canvas environment.
đŦ
AI Chatbot
Students get instant help with coursework from an AI tutor trained on your course materials.
đ
File Management
Upload and organize course materials that power the AI chatbot's knowledge base.
đ
Quiz Generation
Automatically generate quizzes from your course content using AI assistance.
â
Auto-Grading
Save time with AI-powered automatic grading of student submissions.
âąī¸ 12.5 hours spent this week Student engagement data
The Home page gives you an overview of your course. Use the left sidebar
to navigate between Tutorify's features: Chatbot, Files, Students, Quiz, Autograder, and Settings.
7 / 35
Part 2
Tutorify Features
Explore each feature available in the Tutorify sidebar
8 / 35
Feature 1
AI Chatbot
An intelligent teaching assistant powered by your course materials
9 / 35
đŦ
Chatbot
How the AI Chatbot Works
Tutorify uses Retrieval-Augmented Generation (RAG) to answer questions
based on your actual course materials — not generic internet knowledge.
đ
Upload Files
PDFs, DOCX, slides
â
đĸ
Embedding
Convert to vectors
â
đ
Student Query
Ask a question
â
đ
Retrieval
Find relevant passages
â
đ¤
LLM Answer
Generate response
Key Facts
đ¯Answers are grounded in your uploaded course materials, not generic web data
đ§ Supports multiple AI models: GPT-4o, Claude, Gemini, and more
đEach course has its own isolated knowledge base — no data leakage
â°Available 24/7 for students, reducing instructor email load
đInstructors can view chat history and monitor usage
đĄBetter files = better answers. The more relevant and well-structured your uploaded materials, the more accurate the chatbot responses will be.
10 / 35
đŦ
Chatbot
Chatbot Interface
Chat History+ New
đ Search chats...
Stack vs Queue explanation
Binary search complexity
Sorting algorithms review
Recursion examples
Linked list operations
Older
Week 1 - Intro to Python
Variables and types
đ¤AI Teaching AssistantOnline
Model: GPT-4o
đ
How can I help you today?
Choose a prompt or type your own question
đ Explain the concept of recursion
đ Compare arrays vs linked lists
đ Help me study for the midterm
đĄ What is Big O notation?
đ§Š Explain OOP principles
⥠How does a hash table work?
Ask a question about the course...
Send â
đLeft sidebar: View chat history, search past conversations, and start new chats
đRight panel: Suggested prompts for quick access or type your own question
11 / 35
đŦ
Chatbot
Having a Conversation
đ¤ AI Teaching Assistant
Online
đ Hello! I'm your AI teaching assistant for Introduction to CS. How can I help you today?
Can you explain the difference between a stack and a queue?
Great question! Based on your course materials:
A Stack follows Last-In-First-Out (LIFO) — think of a stack of plates. The last item placed on top is the first one removed.
A Queue follows First-In-First-Out (FIFO) — like a line at a store. The first person in line gets served first.
1Type your question in the input box at the bottom
2Click Send or press Enter
3The AI will answer using your course materials
4Ask follow-up questions to go deeper
5Use đ Copy to save responses, or đđ to rate quality
đĄTip: Be specific in your questions for better answers. Instead of "explain sorting", try "explain quicksort with a step-by-step example."
12 / 35
đŦ
Chatbot
Chatbot Features at a Glance
đ
Search History
Search through all past conversations to quickly find previous answers and discussions.
đĄ
Suggested Prompts
Pre-built question templates help students get started with common course topics.
đ
Copy Responses
One-click copy of AI answers to use in notes, assignments, or study materials.
đ
Rate Answers
Thumbs up/down feedback helps improve AI quality and lets instructors monitor accuracy.
đ§
Multiple AI Models
Choose from GPT-4o, Claude, Gemini, and more. Different models for different needs.
đ
Full Chat History
All conversations are saved. Resume any chat where you left off, anytime.
â Instructor Insight: You can view all student chat logs from the Tutorify dashboard to understand what topics students are struggling with, enabling you to adjust your teaching accordingly.
13 / 35
đŦ
Chatbot
Chatbot Tips for Instructors
5 Best Practices
1ī¸âŖUpload comprehensive materials — The chatbot is only as good as the files it can reference. Include lecture notes, textbook excerpts, and slides.
2ī¸âŖUse well-structured documents — PDFs with clear headings and sections produce better search results.
3ī¸âŖMonitor chat logs regularly — Review student questions to identify knowledge gaps and adjust lectures.
4ī¸âŖEncourage students to rate responses — Feedback data helps you identify when the AI gives poor answers.
5ī¸âŖChoose the right AI model — GPT-4o for complex reasoning, lighter models for simple Q&A to save costs.
âPro Tip: Create a "Frequently Asked Questions" document and upload it to the chatbot. This ensures the most common student questions get accurate, consistent answers every time.
â ī¸Important: The AI chatbot is a supplement, not a replacement for instruction. Remind students to verify critical information with course textbooks or by asking the instructor directly.
Supported File Types
PDFDOCXPPTXTXTMD
Upload these formats in the Files section to train the chatbot.
14 / 35
Feature 2
Quiz Generator
Create AI-powered quizzes from your course materials in minutes
15 / 35
đ
Quiz Generator
AI-Powered Quiz Generation
Generate quizzes automatically from your course files using AI. Configure question types,
difficulty, and count — then let the AI create assessments in seconds.
đ
Select File
Choose source
â
âī¸
Configure
Set parameters
â
đ¤
Generate
AI creates quiz
â
â
Review
Edit & export
Capabilities
9
Question Types
3
Difficulty Levels
4+
AI Models
đ¯Questions are derived from your uploaded course files
âī¸Edit, regenerate, or delete any question before exporting
đ¤Export directly to Canvas Quizzes with one click
16 / 35
đ
Quiz Generator
Your Quizzes
The quiz list shows all quizzes you've created. Each card displays the quiz name, difficulty level,
question count, and creation date. Click any quiz to view or edit its questions.
My Quizzes (3)
+ Create Quiz
Midterm Review
Easy
đ 15 questionsđ Multiple Choice, True/Falseđ Created: Feb 10, 2026đ¤ Exported to Canvas
Data Structures Quiz
Medium
đ 10 questionsđ MCQ, Short Answer, Fill-inđ Created: Feb 12, 2026đ¤ Not exported
Algorithms Final
Hard
đ 20 questionsđ MCQ, Essay, Codingđ Created: Feb 15, 2026đ¤ Not exported
đĄTip: Click "+ Create Quiz" to start generating a new quiz. You can create unlimited quizzes and edit them anytime before exporting to Canvas.
17 / 35
đ
Quiz Generator
Create Quiz — Configuration
Quiz Settings
Data Structures Quiz
đ Lecture_Notes_W3.pdfChange
đ§ GPT-4oâž
EasyMediumHard
Multiple Choice5
Short Answer3
Fill in the Blank2
Specifications (Optional)
e.g. "What is the time complexity of binary search?"
Focus on binary trees, hash tables, and graph traversal. Avoid questions about sorting algorithms.
⨠Generate 10 Questions
đĄTip: Use "Additional Instructions" to guide the AI — specify topics to focus on, exclude, or a style to follow.
18 / 35
đ
Quiz Generator
Supported Question Types
Question Type
Description
Best For
Multiple Choice
Four answer options with one correct answer
Factual recall, concept testing
True / False
Statement that is either true or false
Quick knowledge checks
Short Answer
Brief text response (1-3 sentences)
Definitions, brief explanations
Fill in the Blank
Complete a sentence with the missing word or phrase
Terminology, key concepts
Multiple Select
Multiple correct answers from a list of options
Comprehensive understanding
Matching
Match items from two columns
Associations, paired concepts
Essay
Long-form written response
Critical thinking, analysis
Numerical
Answer is a specific number or range
Math, calculations
Ordering
Arrange items in the correct sequence
Processes, chronological events
đĄMix question types for more effective assessments. Combining MCQ with short answer and essay questions tests different cognitive levels (recall, understanding, application, analysis).
19 / 35
đ
Quiz Generator
Generation Progress & Review
Generating...
đ¤
Generating Quiz
AI is creating 10 questions...
7 / 10 questions generated70%
â Q1: Multiple Choice - Created
â Q2: Multiple Choice - Created
â Q3: Short Answer - Created
âŗ Q7: Fill in the Blank - Generating...
Review & Edit
Q1MCQ
âī¸đđī¸
What is the time complexity of searching in a balanced BST?
A) O(1) B) O(log n) â C) O(n) D) O(n log n)
Q2Short Answer
âī¸đđī¸
Explain the difference between a hash map and a tree map.
Q3Fill Blank
âī¸đđī¸
A _______ traversal visits the root node first, then left subtree, then right subtree.
Answer: preorder
Showing 3 of 10 questions · Scroll to see more
20 / 35
đ
Quiz Generator
Export Quiz to Canvas
Once you're satisfied with your quiz, export it directly to Canvas as an assignment. Students will
see it in their Canvas course just like any other quiz.
Export Steps
1Review all generated questions and make any edits
2Click the "Export to Canvas" button
3Set quiz title and point values
4Choose whether to publish immediately or save as draft
5Confirm the export
6Quiz appears in Canvas under Quizzes/Assignments
đ¤ Export to Canvas
This will create a new quiz in your Canvas course with 10 questions.
Data Structures Quiz
10
CancelExport to Canvas
â Success: Once exported, the quiz is fully managed in Canvas. Students take it through Canvas, and grades sync back automatically.
â ī¸Note: Edits made in Canvas after export won't sync back to Tutorify. Make all changes before exporting.
21 / 35
Feature 3
Autograder
AI-powered grading that saves hours of manual work
22 / 35
â
Autograder
How Auto-Grading Works
The Autograder fetches assignments and submissions from Canvas, lets you define rubrics (or generate them with AI),
then uses AI to evaluate each submission — providing scores and detailed feedback.
đ
Select
Choose assignment
â
đĨ
Fetch
Pull from Canvas
â
đ
Rubric
Define or generate
â
đ¤
Evaluate
AI grades each
â
đ¤
Sync
Push to Canvas
đĄRubrics are key. The AI grades based on rubric criteria. More detailed rubrics produce more accurate and consistent grading.
â Grade sync: Scores and feedback are pushed back to Canvas gradebook automatically. Students see their grades in Canvas.
23 / 35
â
Autograder
Select an Assignment
The Autograder shows all assignments and quizzes from your Canvas course. Select one to begin grading.
đĄTip: Assignments are synced from Canvas. If you don't see an assignment, make sure it's been published in Canvas first.
đProgress tracking: Each card shows how many submissions have been received and how many have been graded.
24 / 35
â
Autograder
AI Rubric Generator
Generate with AI
Q1: Explain binary treesâž
20
đ¤ Generate Rubric
Generated Rubric
Excellent (18-20 pts): Complete, accurate explanation with examples. Good (14-17 pts): Mostly correct with minor gaps. Fair (10-13 pts): Partial understanding, some errors. Poor (0-9 pts): Significant misunderstanding or incomplete.
Or Write Your Own
Grade this answer about binary trees on a scale of 0-20:
đĄPro Tip: Keep rubrics concise and specific. Shorter rubrics use fewer AI tokens, reducing costs while maintaining accuracy.
25 / 35
â
Autograder
Grading Student Submissions
Submissions30 total
đ Search students...
AllGradedPending
Jane Smith85/100
Alex Johnson92/100
Sam WilliamsPending
Maria GarciaPending
Chris LeePending
Jane Smith
Submitted: Feb 17, 2026 at 3:42 PM
đ¤ Auto-Gradeđ¤ Sync to Canvas
Q1: Binary TreesQ2: Hash MapsQ3: Graph Theory
A binary tree is a data structure where each node has at most two children. A BST orders elements so that left children are smaller and right children are larger. Common traversals include inorder, preorder, and postorder...
17
/ 20
Good explanation of BST properties and traversals. Missing: real-world applications and time complexity analysis. (-3 pts)
26 / 35
â
Autograder
Auto-Grade vs Manual Grading
đ¤ Auto-Grade
đ¤ Confirm Auto-Grade
AI will evaluate this submission using the rubric and assign a score with feedback.
Student:Jane Smith
Question:Q1: Binary Trees
AI Model:GPT-4o
CancelAuto-Grade
1Click "Auto-Grade" button
2Confirm the AI model and rubric
3AI evaluates and assigns score + feedback
4Review and adjust if needed
â Manual Grading
17
Good understanding of BST structure. Consider adding discussion of balanced vs unbalanced trees and their impact on performance.
Save Score
đĄHybrid approach: Use auto-grade first, then manually adjust scores where needed. This saves the most time while maintaining accuracy.
27 / 35
â
Autograder
Autograder Best Practices
đ
Write Clear Rubrics
Specific criteria lead to consistent AI grading across all submissions.
đ
Spot-Check Results
Review a sample of auto-graded submissions to verify AI accuracy.
đ°
Optimize Token Usage
Keep rubrics concise. Shorter rubrics cost less and often grade better.
đ§
Choose Right Model
Use GPT-4o for essays, lighter models for factual/short answers.
đ¤
Sync Before Deadlines
Push grades to Canvas before students expect them.
đ
Iterate on Rubrics
Refine rubrics based on grading results for improved accuracy over time.
â Time Savings: Instructors report saving 5-10 hours per week using the Autograder for essay and short-answer grading. Focus your time on teaching, not grading.
â ī¸Important: AI grading is an assistive tool, not a replacement for instructor judgment. Always review grades for high-stakes assessments (finals, major projects) before submitting to Canvas.
Recommended Workflow
1. Generate rubrics with AI
2. Auto-grade all submissions
3. Review flagged/low-confidence grades
4. Adjust scores as needed
5. Sync final grades to Canvas
28 / 35
đ
Feature
Managing Course Files
The Files section is where you upload course materials that power the AI chatbot and quiz generator.
Files are processed and embedded for intelligent retrieval.
File Name
Type
Size
Status
Uploaded
Syllabus.pdf
PDF
1.2 MB
Embedded
Feb 1, 2026
Lecture_Notes_W1.pdf
PDF
3.5 MB
Embedded
Feb 3, 2026
Lecture_Notes_W2.pdf
PDF
2.8 MB
Embedded
Feb 5, 2026
Assignment_1.docx
DOCX
845 KB
Processing
Feb 10, 2026
Midterm_Review.pptx
PPTX
5.1 MB
Pending
Feb 12, 2026
How to Upload
1Navigate to Files in the Tutorify sidebar
2Click "Upload File" or drag and drop
3Wait for processing (status changes to Embedded)
đFile Status Guide: Embedded Ready for AI use Processing Being analyzed by AI Pending Queued for processing
â ī¸Supported formats: PDF, DOCX, PPTX, TXT, MD. Max file size: 50 MB. Scanned images (non-OCR PDFs) may not be fully processed.
29 / 35
Part 3
Initial Setup
Connecting Canvas, importing courses, and managing your account
30 / 35
âī¸
Setup
Tutorify Dashboard & Settings
Access the Tutorify Dashboard at demo1.tutorify.ai/dashboard to manage your account, connect Canvas, import courses, and monitor usage.
đ Tutorify Dashboard
đ Dashboard
đ° Budget
đ Billing & Subscription
âī¸ Settings
Welcome back, Professor!
Here is what is happening with your courses today.
Active Courses
3
Total Students
87
AI Queries Today
142
+ Import Course
đ§First-time setup: Go to Settings → Canvas Config, enter your Canvas URL and API Access Token, click "Test Connection", then "Save Settings". After that, use "Import Courses" to bring your Canvas courses into Tutorify.
Students are automatically synced from Canvas LMS. View enrollment status, email, and engagement.
âī¸ Course Settings
Introduction to CS
CS101
GPT-4oâž
On
â Auto-Sync: Students and course information are automatically imported from Canvas. Changes in Canvas (new enrollments, drops) are reflected in Tutorify.
32 / 35
đ°
Management
Subscription & Usage Limits
Tutorify uses a credit-based system. Each AI operation consumes credits based on the model and task type.
Administrators can set usage limits to control costs.
đŦ Chatbot Queries
Per Student
Maximum number of AI chat queries per student per day
65 / 100 used today65%
Default limit: 100 queries/student/day
đ Quiz Generation
Per Course
Maximum questions generated per quiz creation request
40 / 100 questions this month40%
Default limit: 100 questions/course/month
â Autograder
Per Course
Maximum auto-grade evaluations per course per month