MentorOS – AI Student Growth Platform

An AI-powered student growth platform with persistent memory that remembers skills, projects, goals, and learning progress across sessions to deliver personalized guidance and recommendations.

Comment

Inspiration

Many students lose track of their learning progress because existing learning platforms do not remember their long-term journey. We wanted to build an AI-powered student assistant that remembers a student's skills, projects, certificates, resume, and career goals over time. Using Alibaba Cloud Qwen and a persistent memory engine, MentorOS delivers personalized guidance instead of generic AI responses.

What it does

MentorOS is an AI Student Growth Platform with Persistent Memory.

It allows students to:

Register and securely manage their profile Upload resumes for AI-powered analysis Store skills, projects, certificates, and career goals Build long-term semantic memory using ChromaDB Generate personalized learning roadmaps Receive skill-gap analysis Get AI-generated project ideas Track their learning journey over time

Unlike traditional chatbots, MentorOS remembers previous interactions and continuously improves its recommendations.

How we built it

We built MentorOS using:

Alibaba Cloud Qwen (LLM + Embeddings) FastAPI SQLAlchemy PostgreSQL (Neon) ChromaDB JWT Authentication Next.js 15 TypeScript Tailwind CSS Vercel Deployment

The backend stores structured student information in PostgreSQL while semantic memories are stored in ChromaDB. Qwen generates personalized recommendations by retrieving relevant memories before generating responses.

Challenges we ran into

During development we faced several technical challenges:

Designing a persistent memory architecture Combining PostgreSQL with ChromaDB Implementing memory supersession and memory decay Resume parsing and structured information extraction Deploying FastAPI and Next.js on Vercel Configuring environment variables and production database connections Integrating Alibaba Cloud Qwen using its OpenAI-compatible API.

Accomplishments that we're proud of

We are proud that MentorOS:

Successfully integrates Alibaba Cloud Qwen Implements a persistent AI memory system Generates personalized recommendations instead of generic answers Supports resume analysis with structured extraction Maintains long-term learning history Provides a complete full-stack application with frontend and backend deployment

What we learned

This project helped us learn:

Retrieval-Augmented Generation (RAG) Persistent AI memory design Vector embeddings and semantic search Full-stack AI application development FastAPI architecture Next.js application development Production deployment Alibaba Cloud Qwen integration

What's next for MentorOS – AI Student Growth Platform

Our future roadmap includes:

Multi-agent academic assistants University LMS integration Interview preparation assistant Internship recommendation engine AI mentor chat with voice support Mobile application Analytics dashboard for educators Team collaboration and learning communities

Built With

Share this project:

Updates