Seer: Exam Simulation

THE IDEA Hi, Francis here. I'm the lead developer (and founder) of Seer. Seer is a platform dedicated to helping students and anyone who desires to pass their exams do so spectacularly. It does that by attempting to predict which questions you're most likely to see in your exam, then providing a simulated environment to test you on those generated questions. I got the idea for the application back in my first year at university. Back then, whenever I studied past questions for my course, I'd notice that more questions were generated from a particular topic in a course than others. This got me thinking: what if there was a way to predict which questions I was most likely to see in my exams? I knew this was possible based on the amount of past questions I had studied. The Implementation: At first, I decided to upload my course materials to popular LLMs to see if they could spit out questions like those in my past questions, but I quickly realised that doing so was a dead end because LLMs hallucinate and lose context far too quickly. Using my skills as a developer, I decided that I was going to build a platform where students, professionals, and anyone could upload their textbooks, course materials, etc. And they would get questions that would, first, test them on what they've studied so far. And also could technically predict the topics questions would be generated from based on the materials they uploaded. How Seer works: First, you download the app and choose between the student or independent user pathways. If you choose the student pathway, you'll be required to select your school, course, level, and semester, if available. Then, on your dashboard, every subject available in your course would immediately become available. If you go for the independent user path instead, you'd first have to upload your course material yourself, along with things like past questions and a syllabus, if available. After you've completed this step, you can immediately start taking simulated exams consisting of questions generated from the materials you provided. Contact Us I understand that you might have questions you'd like to ask. Please feel free to shoot me a message at: [[email protected]](mailto:[email protected]). Thanks for reading. Francis Obieze