1. Startup Name: Ira Project 2. Startup URL: 3. What year was the company founded? 2024 4. What problem are you looking to solve? In STEM education traditional assignments promote regurgitation of procedures and didn't identify gaps in understanding. Our product aims to give teachers a different way to conduct formative assessments and get an understanding of their students' conceptual knowledge. 5. What is your company going to make to solve this problem? We are building a new type of assignments where students teach concepts to an AI buddy. Ira then "learns" from the student's explanation and attempts to solve a set of questions. Based on their explanations, we are able to build a knowledge graph representing the concepts that the students understood and those they didn't. The teachers can then use these knowledge graphs to adapt their future lesson plans or provide personalized interventions. 6. Is your company incorporated? No 7. What industries are relevant to your company? AI&ML, Edtech 8. Product Demo video URL: 9. Team introduction video: 10. Pitch deck URL: 11. Why is your team a winning team? We have great founder-market fit due to our experience in education. For the past 7 years, Vignesh has been running a non profit which creates after-school learning spaces for marginalised children. He co-founded the non-profit and is actively engaged in teaching. As a PhD student at Georgia Tech, I have taught multiple courses at different universities and am also involved in administration of a private school in Bhubaneswar, India. Vignesh has also been a co-founder thrice before and has successfully scaled products to 100k+ users. I have used neural networks extensively in my research and can build and deploy AI models. My expertise in machine learning and my teaching experience make me the ideal CTO for developing Ira. Both of us are passionate about teaching and technology. We have also known each other since our undergraduate days at Georgia Tech and work well in a team. 12. Please share any traction metrics you have. We have built out an MVP and tested it out with 5 teachers and 5 students. We just finished building out a classroom ready version and are rolling out a pilot to the first school. 13. How do you make money? How much could you make? Currently, we haven't generated any revenue. Our product caters to schools and we will make money by selling to them. The market for K-12 STEM education is massive and we are addressing an underserved problem. 14. Who is your target audience? Our target audience will be K-12 schools in the United States. 15. Who are your competitors? Edtech companies that build AI tools for the classroom or leverage AI to help K-12 students learn. Some examples include FlintK12, Khan Academy, Kahoot. 16. What is your competitive advantage? 1) Most of the products today have students consuming content. By teaching an AI model, we're allowing them to create something - providing a different learning experience. 2) Our product helps students discover what they know and how they know something. 3) We are not creating AI tutors to replace physical teachers. Rather, our product fits into the teacher's workflow and aims to enable them. 17. Why is now the right timing for your startup? The advent of LLMs has made it possible for AI models to understand and reason using natural language. This is the core requirement for creating an AI peer that can communicate and learn STEM concepts from students. 18. What is your customer acquisition strategy? Our first users for this product are going to be the teachers in Vignesh's non profit and Likhit's school. The next few users will be from our personal network that we have built with schools in Atlanta. We have started reaching out to teachers on LinkedIn and Facebook already and we hope to use these channels as well. 19. What are your next major company milestones? To set up a robust distribution channel in the US and build brand awareness for our product. We are also looking forward to the upcoming pilot that will give us valuable feedback on our features.