EduAI – Solution Challenge 2025

Overview

EduAI is an advanced AI-powered learning ecosystem developed by Team Server Overload. Selected as a Top 100 Finalist in the Google Solution Challenge 2025, the platform addresses the disparity in access to quality education. By leveraging Google's Gemini AI and Cloud capabilities, EduAI transforms passive content consumption into an active, personalized learning journey complete with real-time mentorship and adaptive roadmaps.

Recognition

We are proud to be recognized among the top global solutions for our impact on UN Sustainable Development Goal 4 (Quality Education).

Project Demo

Watch our solution in action, detailing the user journey from skill assessment to personalized course generation.

Problem Statement

Uneven Access to Quality Education in the Digital Age: Learners today face content overload, lack of structured direction, and language barriers. High-quality resources are often buried under clickbait, while personalized mentorship remains inaccessible to many students.

Key Features

• AI-Powered Playlist Generation: Automatically aggregates and curates top-tier courses from Coursera, edX, Udemy, and YouTube based on user skill gaps.
• YouTube Quality Analysis: A proprietary AI engine analyzes video comments and engagement metrics to filter out low-quality tutorials, ensuring only the best content reaches the learner.
• AI Mentor Guidance: An intelligent assistant that provides real-time feedback, answers queries, and tracks learning progress.
• Adaptive Skill Assessment: Dynamic quizzes evaluate prior knowledge to tailor content from Beginner to Expert levels.
• Gamified Engagement: Enhances retention through badges, streaks, study groups, and peer discussions.
• Verified Educator Contributions: A dedicated portal for educators to share curated articles, assignments, and courses.

System Architecture

The platform utilizes a microservices-inspired architecture hosted on Google Cloud:

1. User Interaction: Secure authentication via Firebase and Google OAuth 2.0.
2. Orchestration: Cloud Run services manage API requests, routing them to core AI modules.
3. AI Processing: Gemini AI and Natural Language API process user prompts to generate personalized roadmaps.
4. Content Aggregation: YouTube Data API and Custom Search API fetch external learning materials.
5. Storage: User progress and generated content are stored in MongoDB and Google Cloud Storage.

User Interface

Designed for accessibility and focus, the UI adapts to both desktop and mobile workflows.

Performance Benchmarks

We rigorously benchmarked the MVP to ensure a seamless user experience:

• Prompt Processing: ~3.4 seconds (Gemini + NLP processing).
• Playlist Generation: ~40 seconds for a full personalized course.
• Real-Time Mentor Response: ~1.6 seconds.
• Prior Knowledge Assessment Accuracy: ~80% vs. human-assessed baseline.

Technologies Used

• Frontend: Next.js, React.js, Redux, Material UI, Radix UI, Tailwind CSS.
• Backend: Node.js, Express.js, TypeScript, JWT, MongoDB, Firebase.
• AI & APIs: Gemini AI, Google Cloud Natural Language API, YouTube Data API v3, Custom Search API.
• DevOps: Google Cloud Run, Compute Engine, Docker, GitHub Actions, Google Cloud Storage.

Mentorship-Driven Improvements

Based on expert feedback received during the competition, we implemented a Smart Filter Upgrade:

• Granular Filtering: Added filters for Difficulty Level, Content Type, and Source Credibility.
• Impact: Significantly reduced irrelevant search results and improved the alignment of content with user learning goals.

Future Roadmap

• AI-Powered Educator Platform: Allowing teachers to design custom tracks with AI assistance.
• Learning Insights: Analytics to help educators understand regional learning patterns.
• Mobile App: Native Android/iOS support for offline learning.