AI Science Tutor Development

I’m building an intelligent product whose sole purpose is to tutor adult learners in Science. The vision is a conversational, AI-driven experience that diagnoses gaps, adapts in real time, and keeps motivation high through engaging explanations and practice. Here’s how I picture the workflow: a learner signs in, answers a brief diagnostic, then interacts with an adaptive chat interface powered by a fine-tuned large-language model. The back-end tracks progress, surfaces targeted exercises, and serves concise concept refreshers with embedded visuals, simulations, or short quizzes. Think of a personal tutor available 24/7, able to shift from basic principles to advanced applications as the user’s mastery grows. Key elements I need you to handle • Natural-language tutoring engine (Python or Node.js, OpenAI / Hugging Face acceptable) • Adaptive assessment logic that updates a learner profile after every session • Web-based front end—clean, mobile-responsive, preferably React or Vue • Instructor dashboard summarizing cohort and individual performance • Content structure for core Science domains (physics, chemistry, biology) that we can expand later Acceptance criteria 1. A hosted MVP where an adult user can complete a five-minute diagnostic and receive a bespoke tutoring path. 2. Real-time chat capable of answering at least 100 common Science questions with citations. 3. Progress data stored securely; basic analytics visible on the dashboard. 4. Source code, setup instructions, and a short Loom walkthrough at hand-off. If you’ve shipped AI tutoring tools before or have strong NLP plus front-end skills, let’s talk—I’m ready to move quickly from concept to working prototype.