AskGem

AskGem: Your AI mentor for higher studies, personalized questions, university recommendations, and a clear roadmap to academic success.

Comment

AskGem: AI-Powered Academic Roadmaps

Inspiration

Choosing the right university, major, or scholarship can be overwhelming for students. We wanted to create a tool that acts as a personalized academic mentor, guiding students step-by-step with smart, data-driven recommendations. The idea was inspired by how humans mentor students, but with AI scale and consistency.

What it does

AskGem uses Google Gemini 3 to provide students with:

  • Adaptive Q&A to understand their goals and profile.
  • Recommendations for universities, majors, and scholarships.
  • Personalized 12–18 month academic roadmaps. It’s essentially an AI mentor that helps students make informed decisions for higher education.

How we built it

  • Backend: Node.js + Express, structured with controllers, services, and routes.
  • AI Integration: Google Gemini 3 API for dynamic question generation and reasoning-based recommendations.
  • Features: Centralized error handling, sample student profiles for testing, and structured prompt templates for AI responses.
  • Dev Tools: Environment management with .env, modular services, and async error handling for reliability.

Challenges we ran into

  • Designing prompts for Gemini to produce coherent, useful academic advice.
  • Structuring student data to allow the AI to generate meaningful next questions.
  • Handling edge cases where student goals didn’t fit standard paths.
  • Balancing personalization vs. general advice so the roadmap remains realistic and actionable.

Accomplishments that we're proud of

  • Built a production-ready backend capable of generating adaptive AI-driven roadmaps.
  • Successfully integrated Gemini 3 to handle complex student queries.
  • Created a framework that can scale to different academic paths, majors, and scholarship scenarios.

What we learned

  • How to engineer AI prompts for educational guidance.
  • Importance of structured data in generating meaningful AI outputs.
  • Best practices in Node.js/Express for building scalable, maintainable backends.
  • Handling uncertainty in AI responses and creating robust fallback mechanisms.

What's next for AskGem

  • Add a frontend UI for students to interact with the AI directly.
  • Include multi-step roadmap planning with progress tracking.
  • Expand AI capabilities for career guidance post-university.
  • Enable multi-lingual support for global accessibility.

Built With

Share this project:

Updates