PathPilot

Inspiration

Learning today has never been easier, yet it has never been more overwhelming.

Every day, learners are faced with thousands of courses, tutorials, certifications, and roadmaps. The challenge is no longer finding resources—it's knowing which resource to use, when to use it, and what to do next.

Most learning platforms provide static roadmaps or generic recommendations. They rarely adapt to the learner's progress, changing goals, available time, or individual strengths and weaknesses.

We wanted to answer a simple question:

What if every learner had an AI mentor that continuously guided them throughout their learning journey instead of simply giving them another roadmap?

That idea became PathPilot.

What it does

PathPilot is an AI-powered adaptive learning strategist built with Google's Gemini API.

Rather than functioning as another chatbot or course recommender, PathPilot acts as a personalized mentor, recruiter, and career strategist.

After understanding a learner's goals, current skills, study schedule, learning style, and experience, it generates a dynamic roadmap that evolves alongside the learner.

As progress is made, PathPilot recalibrates learning plans, recommends new projects, identifies knowledge gaps, predicts learning risks, and continuously suggests the single highest-impact action the learner should take next.

The experience is presented as a guided multi-stage journey consisting of:

Discover – Personalized opportunity analysis Evaluate – Recruiter-style career assessment Coach – AI mentor feedback Measure – Career intelligence dashboard Protect – Learning risk analysis Launch – Personalized next mission

Instead of overwhelming users with information, each stage reveals focused insights that help learners make better decisions with confidence.

How we built it

PathPilot was designed as a modern full-stack AI application.

Our technology stack includes:

React for the interactive frontend Vite for fast development Express.js for the backend Google Gemini API through Google AI Studio for AI reasoning Local Storage for client-side persistence Responsive UI with modern animations and dashboard components

One of our biggest architectural decisions was separating the AI's identity from its tasks.

Rather than relying on one enormous prompt, PathPilot uses a structured system prompt to define its coaching behavior while sending focused task prompts for roadmap generation, progress analysis, mentoring, recruiter reviews, and opportunity analysis.

This modular prompt design produces faster, more focused, and more maintainable AI responses.

Challenges we ran into

The biggest challenge wasn't building another AI chatbot—it was designing an AI experience that genuinely feels like a mentor.

Early versions produced long responses filled with valuable information, but presenting everything on one page made the experience overwhelming.

To solve this, we redesigned PathPilot into a guided consultation where learners progressively move through different stages of career development.

Another challenge was balancing personalization with performance. Rich AI responses can become slow and repetitive if prompts are too broad, so we optimized our prompting strategy by separating permanent system instructions from specialized task prompts and caching AI-generated insights during navigation.

This significantly improved both responsiveness and user experience.

Accomplishments that we're proud of

What we learned

Building PathPilot taught us that successful AI products are not defined solely by the intelligence of the model—they are defined by the experience surrounding it.

We learned the importance of:

Designing AI around real user workflows rather than chat conversations. Using structured prompt engineering instead of one massive prompt. Breaking complex AI outputs into digestible experiences. Combining software engineering principles with AI reasoning to create products that feel intentional and practical.

Perhaps the most important lesson was that AI should empower decision-making, not replace it.

PathPilot doesn't simply tell learners what to study—it helps them understand why that recommendation matters and what opportunities it unlocks.

What's next for PathPilot

PathPilot is designed to grow beyond this hackathon.

Future enhancements include:

GitHub portfolio analysis Resume and CV reviews Mock technical interviews Calendar integration Learning streak analytics AI-generated interview preparation Team learning support Real-time job market insights Mentor collaboration features

Our vision is for PathPilot to become more than a roadmap generator.

We envision it as an intelligent career companion that grows with learners from their first lesson to their first job—and beyond.

Built With

• gemini-api
• google-ai-studio
• prompt-engineering