Visual-GPT

Inspiration

As a mathematics major at the University of Melbourne, I've seen firsthand the silent crisis unfolding on our campuses. I've spent hundreds of hours volunteering for bridging programs designed to help first-year STEM students succeed, and through this, I've built deep relationships with both the students who feel lost and the dedicated lecturers struggling to teach classes with such diverse knowledge levels. The inspiration for Visual-GPT comes directly from their stories and their daily pain points – it's a problem I haven't just studied, but one I have lived from every angle.

What it does

Visual-GPT is an interactive AI learning companion designed to fix the most expensive leak in our education system. It tackles the "knowledge gap" that causes 1 in 5 Australian university students to drop out, a problem that costs the economy billions.[1] We transform dense textbook material into engaging, bite-sized videos. Our core innovation, Visual-GPT, allows students to pause any video and ask a question in natural language. The platform then instantly generates a custom micro-lecture with new visuals to answer that specific question. Following this, it creates an adaptive learning roadmap, empowering students to review foundational concepts or proceed with university-level content. Finally, it connects students who have asked similar questions into peer support forums, fostering a collaborative community to combat the isolation that often leads to dropping out.

How I built it

I identified the pain points on the front lines, and my co-founder, Jasper, our technical engine, built the solution. He has developed a proprietary method to generate our core animated content with unmatched speed and accuracy. This allows us to be incredibly responsive to user needs. Our process is simple and rapid: I identify the challenges students and educators face, and Jasper builds the technological solution, often overnight. Our platform is built on a foundation of user-centric design, informed by extensive conversations with students, university staff, and high school teachers.

Challenges I ran into

A primary challenge has been ensuring that our AI-generated content is not only accurate but also pedagogically sound and truly engaging. It's not enough to simply convert text to video; the content needs to simplify complex concepts effectively. Another hurdle has been developing the AI to accurately assess a student's knowledge level based on their questions and provide truly personalized follow-up recommendations. Furthermore, creating a seamless and intuitive user experience that encourages active participation rather than passive viewing has been a continuous process of iteration and refinement based on user feedback.

Accomplishments that I'm proud of

Our greatest accomplishment is the overwhelmingly positive validation we've received from our target users. Students have told us that a visual, interactive tool they can ask questions to would be a "game-changer." University of Melbourne teaching staff have confirmed that bridging the knowledge gap is a "persistent challenge" and see our platform as a powerful solution. We've also garnered enthusiasm from high school teachers who see Visual-GPT as the perfect tool to prepare students from all backgrounds for the rigors of university STEM. This broad-based support from students, university staff, and high school teachers validates that we are addressing a real and significant need in the education system.

What I learned

Through this process, I've learned that the most effective solutions come from a deep and personal connection to the problem. My experience on the front lines of student support at the University of Melbourne has been invaluable in shaping a product that resonates with the real needs of students and educators. I've also learned the power of a symbiotic founder relationship; my deep market insight combined with my co-founder's rapid and brilliant technical execution allows us to innovate at an incredible pace. Finally, the feedback from our target users has reinforced the importance of building with them, not just for them.

What's next for Visual-GPT

Our ambitious vision is to become the leading platform for visual and interactive STEM education globally. Our immediate next steps are to launch our freemium B2C model to drive individual student adoption. Concurrently, we will be pursuing a B2B strategy, offering annual institutional packages to high school STEM departments and university teaching and learning units. We see a clear path to scalability, expanding our platform's application from primary and high school to universities, both in Australia and eventually, worldwide. We are confident that Visual-GPT has the potential to not just be a successful business, but to fundamentally transform STEM education for the better.

pitch: https://youtu.be/itc3mhEMKKU demo video: https://youtu.be/itc3mhEMKKU

Built With

• ai
• api