Learner course marketplace featuring category filters, instructor profiles, diamond pricing and direct contact with seller.
Membership comparison between Standard Free and 150-Diamond VIP tiers
Diamond currency top-up modal supporting 14 payment rails and transparent conversion rates.
Agentic support pipeline demonstrating orchestrator intent routing and domain guardrails.
Learner's unlocked course library updated immediately after diamond balance deduction.
Adaptive diagnostic assessment launch screen for personalized student skill benchmarking.
Interactive assessment interface evaluating conceptual mastery and test performance.
Automated diagnostic results pinpointing knowledge gaps and prescribing a study plan that will send result to gmail.
Hub for course management, content authoring, and sales overview.
Today's sales analytics tracking across all published courses.
Real-time student diagnostic cards displaying quiz scores and AI-detected concept weaknesses.
Interactive growth analytics showing income trajectories and trends per course over time.
Direct communication inbox facilitating real-time learner inquiries and seller support.
Seller payout modal allowing earnings withdrawal across multiple local payment methods.
Autonomous course generator synthesizing complete module syllabus and interactive quiz questions.

Inspiration

Selling courses online today is painful for many creators. They have to post constantly on social media and manually message bank account numbers back and forth to get paid, which is messy and unsecure. At the same time, learners using basic AI chatbots quickly realize that these tools forget past conversations and cannot give deep personalized help. AdaptLearn solves both problems. It gives creators a single marketplace hub where buyers pay using a simple Diamond wallet with local eWallets, while offering learners a personal AI tutor that runs 100% locally on their computer.

What it does

AdaptLearn is a learning and skill marketplace for everyone. Whether someone wants to learn new knowledge, personal finance, practical life skills, or tech topics. They can browse courses and unlock them using platform Diamonds. The platform includes:

A Learner Hub: Browse courses, take smart quizzes, chat directly with course sellers, and receive lesson summaries sent to your email inbox.
A Seller Hub: Create courses, set prices in diamonds, track sales analytics, and withdraw earnings.
7 AI Helper Agents: Specialized agents (Router, Tutor, Examiner, Study Planner, Email Sender, Support, and Seller Copilot) that answer questions, grade quizzes, write study plans, and help sellers generate course outlines. ## How to built it Built AdaptLearn using a modern full-stack setup:
Frontend: Built with React 18, Vite, Tailwind CSS, Radix UI, and Recharts for analytics graphs.
Backend: Built with Node.js, Express, and TypeScript.
Database: Managed with SQLite and Prisma ORM to store user accounts, diamond balances, chat sessions, and long-term learning history.
Local AI Engine: Powered by Llama 3.2 running locally on the computer using Ollama. Our custom TypeScript service routes user messages to the right AI helper and compresses long chat logs so conversations stay fast.
Email Service: Built with Nodemailer to deliver lesson summaries to users.

Challenges when ran into

One big challenge was connecting local Ollama streaming responses smoothly to the React frontend so users get instant answers. Another challenge was preventing the AI from slowing down during long conversations. Solved this by creating an automatic chat compressor that summarizes older messages into short summaries and also worked hard to make sure the AI accurately picks the right helper agent for every user message.

Accomplishments that have proud of

Proud of building a working 7-agent AI system that runs completely locally on standard computers with zero monthly API bills. Proud of creating a smooth Diamond wallet system that supports local payment methods and allows sellers to withdraw their earnings easily.

What learned

Learned about how to design and route multi-agent AI systems locally using TypeScript and Ollama. We also learned how to manage long-term database memory for AI helpers and how to build a clean marketplace experience for both buyers and sellers.

What's next for

Next, add monthly subscription options for creators and a referral reward system where learners earn bonus diamonds for inviting friends. Support live interactive workshops where AI helpers assist hosts in real-time Q&A.