PERSONAL PATH FINDER

Inspiration

We were inspired by the challenges students face when preparing for technical careers. Many learners follow generic roadmaps, watch countless tutorials, and practice coding without knowing their actual skill gaps or whether they are ready for industry requirements. We wanted to create a smarter learning platform that understands each learner's current skills and guides them toward their career goals through a personalized, practical, and adaptive learning experience.

What it does

PathAI is an AI-powered personalized career pathfinder that helps learners develop the skills required for their target technical roles.

Conversational Skill Assessment: Identifies career goals, existing skills, experience levels, and learning preferences through an AI-powered diagnostic conversation.
Adaptive Learning Roadmaps: Generates personalized learning paths using a prerequisite graph, prioritizing missing skills and skipping topics the learner has already mastered.
Ranked Learning Resources: Recommends relevant learning content based on the learner's skill gaps and career objectives.
Video and Coding Workspace: Combines video-based learning with hands-on coding challenges in a unified environment.
Sandboxed Code Execution: Evaluates Python and JavaScript code against public and hidden test cases, providing practical feedback on coding performance.
AI-Powered Mentorship: Helps learners understand concepts, troubleshoot problems, and overcome learning obstacles.

Our goal is to make technical learning more focused, interactive, and aligned with real-world job requirements.

How we built it

We built PathAI using a modern full-stack architecture that integrates frontend development, backend APIs, AI-powered personalization, database management, and code execution.

Frontend: React, Vite, TypeScript, and Tailwind CSS for a responsive and interactive user interface.
Backend: FastAPI with Python for core application APIs, alongside Express.js and Node.js for server-side rendering and sandbox execution.
AI Integration: LangChain and OpenAI GPT-4o / GPT-4.1 Nano for conversational diagnostics, structured responses, and personalized learning recommendations.
Database: Supabase and PostgreSQL for structured data storage, supported by Row Level Security for access control.
Roadmap Visualization: React Flow and Dagre for displaying personalized learning paths as prerequisite graphs.
Code Execution: Server-side subprocess execution for Python and JavaScript, with timeouts and environment isolation to support safer code evaluation.

The platform connects these components into a continuous learning workflow: diagnose skills, generate an adaptive roadmap, discover relevant content, practice coding, and receive contextual guidance.

Challenges we ran into

One major challenge was designing a personalized roadmap that adapts to different skill levels instead of forcing every learner through the same sequence. We addressed this through a prerequisite graph that represents dependencies between topics and helps prioritize important skill gaps.

Another challenge was converting conversational inputs into structured information that the system could use reliably. AI responses needed to follow predictable formats so that career goals, skills, and learning preferences could be processed consistently.

Integrating learning resources with hands-on coding practice was also challenging. We designed a two-mode workspace to make switching between video learning and coding exercises more seamless.

Finally, executing user-submitted code required careful attention to timeouts, process isolation, and hidden test cases. These challenges highlighted the importance of backend validation and security when building interactive learning platforms.

Accomplishments that we're proud of

We are proud of designing a unified learning experience that connects career planning, personalized education, coding practice, and AI mentorship in one platform.

Our key accomplishments include:

Designing a conversational approach to identify learners' skills and career objectives.
Creating an adaptive prerequisite-graph concept for personalized learning roadmaps.
Combining ranked learning resources with hands-on coding challenges.
Planning real server-side code evaluation using public and hidden test cases.
Building a modular architecture that separates onboarding, roadmap visualization, content transformation, and AI mentorship.

Most importantly, PathAI focuses on validating practical skills rather than relying only on completed courses or watched tutorials.

What we learned

Through this project, we learned how to bring together frontend development, backend APIs, AI integration, database design, graph-based visualization, and code execution into a single application.

We also learned that effective AI applications require more than generating text. They need structured outputs, reliable data flow, meaningful evaluation, and clear feedback for users.

Designing adaptive learning paths helped us understand the importance of prerequisites and skill dependencies. Working on sandboxed execution also reinforced the need to consider security, resource limits, and backend validation from the beginning.

Most importantly, we learned to approach technology from the user's perspective: a successful learning platform should not simply provide more content; it should help learners identify what matters and make measurable progress.

What's next for PathAI

We plan to extend PathAI with features that make personalized technical learning even more effective:

Progress Tracking: Monitor completed skills, coding accuracy, and roadmap progress over time.
Smarter Adaptation: Update learning paths dynamically based on assessment results and coding performance.
Expanded Language Support: Add more programming languages and improve code evaluation capabilities.
Career Benchmarking: Map learner skills to job descriptions and industry competency requirements.
Learning Analytics: Provide dashboards showing strengths, weaknesses, and areas requiring improvement.
Enhanced AI Mentorship: Deliver more contextual explanations, debugging guidance, and targeted practice questions.
Secure, Scalable Execution: Strengthen sandbox isolation and deployment infrastructure as the platform grows.

Our long-term vision is to make PathAI a reliable companion for learners who want to move from fragmented learning experiences to structured, personalized, and demonstrable career readiness.