About PQ-ACE

PQ-ACE is our AI-powered study companion that brings past exam questions into one place—so students can practice smarter, not harder.

Inspiration

We’re a team of four who’ve all felt the stress of hunting down old tests and notes at 2 AM before an exam.

• Scattered resources: Between friends’ hand-written notes, PDF dumps, and random websites, studying felt more like treasure hunting than learning.
  
• Anxiety spikes: We saw classmates averaging only 60% on tests and reported big anxiety gaps because they couldn’t find the right practice material.
  
• “What if AI could…?” A casual chat about Google’s new Gemini model turned into: “Let’s build a chatbot that ingests past questions and quizzes us back!”

What We Learned

Building PQ-ACE was as much a learning journey as it was a coding sprint.

• Full-stack fluency: We deepened our React skills and got hands-on with Firebase Auth, Firestore, and Storage.
  
• AI integration: Wrangling the Google Generative AI API taught us real-world prompt design, file encoding, and streaming responses.
  
• Design & UX: Tailwind CSS + Vite gave us a crash course in responsive, theme-toggle UIs—and the power of utility-first styling.
  
• Team workflows: We started with VS Code Live Share, hit latency walls, then pivoted to Git + GitHub branches—boosting our productivity by 50%.

How We Built It

1. Frontend (React + Vite):
   
   • AuthContext.jsx for Google Sign-In
     
   • FilterBar, PastQuestionCard, SavedQuestionsPage components
     
   • GeminiChatPage.jsx: upload documents → base64 → AI chat UI
     
2. Backend & Data (Firebase):
   
   • Authentication (Google)
     
   • Firestore collections: users, pastQuestions, savedQuestions
     
   • Storage for PDFs/images
     
3. AI & Vision:
   
   • Google Gemini 1.5 Pro for question generation
     
   • Cloud Vision (OCR) to parse image-based questions
     
4. Styling & Build:
   
   • Tailwind CSS for dark/light themes and responsive layouts
     
   • Vite dev server for instant reloads
     
5. Collaboration:
   
   • GitHub branches + pull requests
     
   • Code reviews & peer pairing for bug hunts

Challenges Faced

• Live Share latency: Real-time editing was laggy, so we switched to Git workflows mid-hackathon.
  
• AI setup delays: Configuring credentials, CORS rules, and file-upload streams cost us nearly 2 extra days.
  
• Time & sleep debt: Late-night sprints taught us why realistic sprint planning (and coffee budgets) matter!
  
• Edge cases & errors: Handling upload failures, incremental loading of large PDFs, and stream errors in the chat UI pushed us to build robust retry logic.

What’s Next for PQ-ACE

1. Expand AI Features
   
   • Personalized study recommendations
     
   • Automated content generation
     
2. LMS Integration
   
   • Connect with popular Learning Management Systems for seamless resource access
     
3. Mobile App Development
   
   • Native apps for iOS and Android
     
4. Community Building
   
   • Forums, user profiles, and social sharing features
     
5. Wider Language Support
   
   • Cater to a global student community
     
6. Continuous Improvement
   
   • Gather user feedback and iterate on the platform

We believe that PQ-ACE has the potential to transform the way students study and collaborate, and we are excited to continue developing and improving the platform.

Built With

• css
• firebase
• geminiapi
• javascript
• markdown
• node.js
• react
• react.js
• tailwind
• ui
• webspeechapi