Our team created AskCanvas, a solution that aims to connect the gap between students and teachers turning unanswered questions and confusion into class-level information for instructors. Together, as team 15, we combined brainstorming, professor interviews and individual research into a feasible idea that addresses everyday problems in education. As I come from an International Relations background, I perceived AI mostly as an uncertain, slightly intimidating tool, or just a simple study aid. Building AskCanvas with my team changed how I understand and view it. The biggest lesson I learned from this project is that AI stopped being a game changer on its own, but that its value comes from what it is combined with: people, context and a real problem.
A Problem Closer to Home
Every painting starts with a rough sketch, and the early stages of AskCanvas were no different. We worked entirely together, filling in each other’s gaps and challenging each other’s thinking. The initial conversations had our team’s focus shift from AI solutions for private businesses to a problem closer to home, rooted in our daily lives as university students. To sharpen the problem, I sent short interviews to professors and former supervisors through my university contacts. Initially, I pictured the problem mostly from the student’s side. Shyness and embarrassment stopping students from asking, courses being evaluated only after exams, when it is already too late to change anything. Professors’ feedback shifted our focus to instructors, flooded with questions across multiple channels with limited hours and budget. In the end, we decided to blend both perspectives and targeted the gap between student confusion and what instructors can realistically do. Surprisingly, in the process, I discovered that my former professor was already working on the same problem, which gave us the ultimate validation and motivation to move forward with our solution, rather than discouraging us. I greatly enjoyed our team efforts put into this project, mostly because the idea we ended up with was a direct solution to the struggles both we as students and our instructors actually find challenging.
Beyond the Chatbot
Once the outline was in place, we divided the workload into individual sections. Ecosystem, defensibility and future market trends matched my interests, as I am genuinely curious about where markets are heading and truly believed our idea could hold its ground in the future. Mapping the ecosystem showed me that Canvas owns the position where course data, instructors and students already meet, but not the AI itself, for which it depends on LLM providers. This discovery shaped my defensibility argument, that access to AI is no longer an advantage on its own, what you combine it with is. Working on an AI-dominated project made us realize that AskCanvas, like any other innovative idea, is defensible, but not unbeatable, since competing platforms hold the same position with their own clients. In my opinion, admitting this openly made our argument even stronger. Looking back, the hardest part of my research was finding sources on future market trends. As the field moves faster than academic research can keep up, I had to critically evaluate and check each and every prediction and distinguish evidence from just guesses. This in the context of AI felt like aiming at a moving target.
Questions from the audience
Early in the process, our lecturer pointed out that our assumption that “professors want more insight” was not enough, and challenged us to prove that AI is actually able to give instructors better information than reviewing questions manually. We tested it and found that AI reduced 30 raw student questions to only 13 items to review. Our efforts on building AskCanvas were displayed at the Demo Day, which resulted in receiving surprisingly good feedback. Some fellow students mentioned they wished AskCanvas had existed the moment they started university, which fuelled my motivation and growing interest in AI development. There were doubts too. Questions such as “How would you monetize it?” and “Why not just ask ChatGPT?” gave us an opportunity to rethink our solutions once more and put our defensibility argument to a live test. Adding the fee to the Canvas licence universities already pay for resolved the first dilemma. For the second, our strategy on AskCanvas having access to all compiled course data as well as to a real instructor set it apart from standard AI tools we all use.
Strength in Numbers
In a lecture hall, one student’s question often speaks for many; in our team, one idea often sparked the next. We used shared documents combined with individual research documents and met as regularly as our schedules allowed. Different members took the lead in different situations, but we always supported each other and reached decisions through a consensus. With two team members coming from an international relations background, including me, and two from business, we were able to tackle the problem from various angles. The most rewarding feeling was turning four people’s research into one coherent story.
What Stays after the Project
This project substantially changed how I see AI. I now can see its strategic value, usefulness and how it can be used in almost every field. Now that everyone has access to it, AI is no longer a stand-alone game changer, but it is most successful when combined with a human factor or an inspiring idea. AskCanvas reflects this by implementing AI but always leaving the final decision to the instructor. Besides this, I learned that, from a business strategy angle, the key question is no longer whether to use AI, but what to combine it with that competitors cannot easily copy. That is precisely the leverage our team aimed for with AskCanvas, turning widely available AI into something only Canvas can offer. Applying abstract lecture concepts to our own project taught me more than any textbook could. If I could, I would put more time into this project, because the whole journey showed me that the best form of learning is learning by doing.
