Replit-Built AI Training Platform for AMC/AIME

Description: Built MathGauss as an AI-powered training platform for students preparing for math competitions like AMC and AIME, developed and iterated entirely in Replit. The platform delivers contest-style problems, instant AI explanations, and adaptive practice sessions so students can sharpen speed, accuracy, and problem-solving skills in a focused environment.

Designed a problem engine that serves curated AMC/AIME-style questions with tags for topic and difficulty, plus timed practice modes.

Integrated an AI tutor that gives step-by-step hints and full solutions, and recommends follow-up problems based on recent mistakes.

Implemented clean math rendering with MathJax so LaTeX-style formulas, equations, and multi-line derivations display correctly on all devices.

Built a responsive dashboard where students can track accuracy, average solve time, topic-wise performance, and recent sessions.

Added streaks, session goals, and progress summaries to keep learners motivated over long preparation periods.

Optimized delivery with CDN, HTTP/3, and Google Cloud load balancing so practice remains smooth even during peak usage.

Used Replit as the main dev environment for rapid prototyping, live previews, and quick deployments to production.

Tech Stack: Replit · Node.js · Express · Radix UI · shadcn/ui · Tailwind CSS · MathJax 3.2.2 · Lucide · Polyfill · Open Graph · HTTP/3 · jsDelivr · Google Cloud (IaaS, Cloud CDN, Cloud Load Balancing, Cloud Trace)