Inspiration
As CUNY students, we know how overwhelming it can be to receive a long syllabus at the beginning of a semester. Important exam dates, assignments, grading policies, and course topics can be buried across 10 or more pages of information.
We wanted to solve a bigger problem than just making the syllabus shorter. Even after reading a syllabus, students may still ask themselves: "What should I study first?"
This inspired us to create Course Canvas, an AI-powered tool that turns a complicated syllabus into a visual course dashboard and creates personalized learning roadmaps. Our goal was to make studying feel less overwhelming by giving students a clear path from where they are now to where they need to be.
What it does
Course Canvas allows students to upload their syllabus or course materials and uses Gemini to understand and organize the information.
It can identify important information such as:
Assignment and exam deadlines Course topics and learning objectives Grading percentages Required readings and materials Important course policies
The main feature is our AI Study Roadmap. Students can enter a goal, such as "I need to learn standard deviation," and Course Canvas determines what concepts they should understand first.
Instead of jumping directly into standard deviation, the roadmap can guide the student through prerequisite concepts such as data sets, mean, variance, and then standard deviation.
Course Canvas can also create a personalized study schedule based on the student's available time. If a student has only a few hours available during the week, the AI can break the learning roadmap into manageable study sessions.
Our goal is to answer three questions for every student:
What do I need to learn?
What do I need to know first?
How do I get there?
How we built it
We built Course Canvas as a web application with an interactive frontend and an AI-powered backend.
Students can upload their course materials through the frontend. The backend sends the relevant content to Gemini, which analyzes the document and extracts structured information about the course.
We then use that information to create the course dashboard, deadlines, topic breakdowns, and personalized study roadmaps.
The AI roadmap works by identifying relationships between concepts. When a student chooses a target topic, Gemini analyzes the topic and generates the prerequisite concepts that lead toward it.
We also designed the system so that the student's available study time can be taken into account when generating a schedule.
Gemini was especially useful because it can work with long documents and different types of course materials, allowing us to turn information that is normally static and difficult to navigate into something interactive.
Challenges we ran into
One of our biggest challenges was making the AI-generated study roadmaps actually useful.
It is easy to ask an AI to generate a study plan, but a generic plan does not necessarily account for what a student already knows or what they need to learn first. We had to think carefully about how to prompt Gemini to identify prerequisite knowledge instead of simply listing topics.
Another challenge was dealing with different syllabus formats. Every professor structures their syllabus differently. Important information might be in a table, paragraph, image, or spread across multiple pages. We needed our system to handle this variety while still extracting the information students actually need.
We also had to balance our ideas with the time available to build the project. There are many features we wanted to include, so we focused on creating a strong core experience rather than trying to build everything at once.
Accomplishments that we're proud of
We are most proud of turning a static syllabus into something that can actually guide a student's learning.
The AI Study Roadmap is especially important to us because it goes beyond summarization. Instead of simply telling a student what a topic means, Course Canvas helps answer what they should understand before learning that topic.
We are also proud of combining multiple pieces of information — course deadlines, learning topics, prerequisites, and available study time — into one experience.
Most importantly, we built something that addresses a problem we have personally seen among college students. Course Canvas is not just an AI demo; it is designed around a real student workflow.
What we learned
We learned that building an AI-powered application requires much more than connecting an API.
We learned how important it is to structure information clearly, design effective prompts, handle unstructured documents, and think about how AI-generated results should be presented to users.
We also learned that personalization is one of the most valuable parts of AI. A generic explanation can help anyone, but a system that understands a student's goal, prerequisites, deadlines, and available time can provide much more useful guidance.
Most importantly, we learned to think about education as a connected path rather than a list of subjects. Understanding what comes before a concept can be just as important as understanding the concept itself.
What's next for Course Canvas
There is a lot we would like to add to Course Canvas in the future.
We want to make the study roadmap more interactive so students can mark concepts as completed and have the roadmap automatically adapt based on their progress.
We also want to add:
AI-generated flashcards from course materials Practice quizzes based on upcoming exams Progress tracking and mastery levels Calendar integration Smarter scheduling when deadlines change Support for multiple courses at once Personalized explanations based on a student's knowledge level Professor and course-specific learning resources
Our long-term vision is for Course Canvas to become a personal academic navigator. Instead of students simply receiving information from their courses, Course Canvas would help them understand what that information means, what they need to learn next, and how to reach their academic goals.
Log in or sign up for Devpost to join the conversation.