SkillQuest AI

• 

• 

• 

• 

• 

• 

Inspiration

Learning online today is overwhelming. When someone wants to learn a new skill—whether it’s data analytics, cloud computing, or product management—they’re usually faced with thousands of videos, courses, and articles. The hardest part isn’t finding content, it’s figuring out what to learn next.

Most learning platforms provide large libraries of courses but don’t truly adapt to the learner’s background, goals, or progress. We wanted to build something different: a system that acts like a personal learning guide that walks you step-by-step through the skills you actually need.

Our goal was to create a platform that removes the noise from online learning and answers one simple question:

“What should I work on next?”

What it does

SkillQuest is an AI-powered personalized learning platform that helps professionals focus on the right skills at the right time.

When users sign up, they go through a short onboarding process where they provide their job role, current skill levels, learning goals, and preferences. The platform then creates a scoped learning environment based on the user’s selected domain and skills.

From there, the system:

Builds a dynamic skill map showing current mastery and progress

Detects weak spots based on quizzes, engagement signals, and performance

Generates personalized learning recommendations explaining why a skill is important now

Provides targeted learning resources including YouTube snippets and curated tutorials

Creates realistic practice scenarios tailored to the user’s background

Tracks progress, streaks, and XP to keep learners motivated

Instead of browsing endless tutorials, users get a clear, focused learning path designed specifically for them.

How we built it

SkillQuest was built as a full-stack AI-powered learning system.

The frontend was built with Next.js and React, creating a responsive dashboard with onboarding flows, skill maps, recommendations, analytics, and a practice lab.

On the backend we used PostgreSQL with Prisma to manage user profiles, skill states, learning history, and recommendations. Redis was integrated to cache recommendation results and improve performance.

Authentication is handled using Firebase Authentication integrated with NextAuth, allowing secure login and session management while linking Firebase identities with PostgreSQL user records.

To power intelligent learning features, we implemented several learning models:

A mastery model that calculates skill proficiency based on quiz accuracy, speed, and practice results

A weak-spot detection model that identifies struggling concepts using behavioral signals

A recommendation engine that ranks the most relevant next skills based on goals, skill gaps, and engagement

A spaced repetition algorithm to determine when concepts should be reviewed

An AI scenario generator that creates realistic practice tasks using OpenAI

We also integrated the YouTube Data API to fetch relevant video snippets and timestamps that explain specific concepts quickly.

All learning content is scoped to the user’s selected domain so recommendations remain focused and relevant.

Challenges we ran into

One of the biggest challenges was designing a system that produces meaningful personalized recommendations instead of generic suggestions.

Learning platforms often rely on simple keyword matching, but we wanted recommendations to consider multiple signals such as skill mastery, weak spots, learning goals, and engagement behavior.

Another challenge was implementing a scoped learning architecture so users only see content relevant to their chosen domain and skills. Without this filtering, recommendations could quickly become noisy and confusing.

We also encountered technical challenges while integrating multiple systems together, including authentication with Firebase and NextAuth, handling database constraints during onboarding diagnostics, and fixing state management issues in the quiz timer.

Ensuring that recommendations, practice scenarios, and resources all stayed synchronized required careful backend architecture.

Accomplishments that we're proud of

We’re proud that we built a working prototype of SkillQuest, not just another course platform.

Some accomplishments include:

Designing a scoped learning architecture that keeps recommendations focused on a user’s domain and selected skills

Building a dynamic skill map that visualizes mastery and learning progress

Implementing AI-generated practice scenarios tailored to a user’s background

Creating weak-spot detection using behavioral learning signals

Integrating YouTube snippet recommendations for fast concept learning

Developing a full learning dashboard with analytics, XP progression, and achievements

Most importantly, the system helps users answer the question “What should I learn next?” in a clear and personalized way.

What we learned

Working on SkillQuest taught us a lot about building AI-powered learning systems.

We learned how to combine multiple data signals such as quizzes, engagement behavior, and goals to generate meaningful recommendations. We also gained experience designing scalable backend systems that integrate AI services with real application logic.

Another key lesson was that building an effective AI product is not just about using machine learning models, but about designing the right user experience around intelligent recommendations.

We also learned the importance of keeping learning focused and reducing cognitive overload for users.

What's next for NextLearn AI

This project is just the beginning of what SkillQuest could become.

Our next steps include:

Expanding the platform to support more professional domains and skills

Adding AI-generated micro-lessons for weak concepts

Introducing adaptive quizzes that automatically adjust difficulty

Implementing peer learning and social skill comparisons

Adding voice-based AI learning assistants

Improving the recommendation engine with more behavioral data and feedback loops

Our long-term vision is to build a platform that becomes the default way professionals continuously learn and upgrade their skills throughout their careers.

Built With

• and-app-state-storage.-we-integrated-the-youtube-data-api-to-surface-learning-video-resources
• and-designed-the-ai-layer-around-gemini-powered-course-and-recommendation-generation-for-personalized-learning-paths
• and-gamified-tasks.-the-app-also-uses-route-based-game-pages
• authentication
• gemini
• next.js
• react
• reusable-component-architecture
• skill-breakdowns
• supabase
• tailwind
• we-used-supabase-for-the-database
• with-a-react/next.js-style-frontend-and-tailwind-css-for-the-ui.-for-backend-ready-architecture-and-data-persistence
• youtubeapi