🤖 EngineerlyAI

A web-based application designed to provide intelligent tutoring using course-specific textbooks and prompts. It consists of a Next.js frontend deployed on Vercel and a Flask backend deployed on Google Cloud Run. The application leverages Google Generative AI (Gemini) for generating responses based on user chosen course and prompt.

100 Page Performance Score

✨ Features

• Course-Specific Prompting: Select a course using a /course_code prefix to tailor the AI's response.
• AI-Generated Answers: Provides detailed and accurate answers to user queries using textbooks and the Gemini model.
• Dynamic Dropdowns: Helps users quickly select available course codes while typing.
• Responsive Chat History: Displays user and AI responses in a clean chat-like interface.
• Automatic Textbook Management: Uploads textbooks to Gemini periodically for the most up-to-date learning resources.
• Smart Course Inference: Detects course textbook to reference based on the first prompt in a chat to reduce prompt redundancy.

🛠️ Tech Stack

🌐 Frontend

• Framework: Next.js
• Deployment: Vercel
• Key Features:
  • Dynamic dropdowns for course selection.
  • Real-time chat interface with loading indicators for bot responses.
  • Auto-complete and shortcut support for commands.

🖥️ Backend

• Framework: Flask
• Deployment: Google Cloud Run
• Generative AI: Google Generative AI
• Scheduler: Uses apscheduler to automate textbook uploads.

🚀 Key Functionalities

🌐 Frontend

• Prompt Input:
  • Use a /course_code prefix to select a course.
  • Autocomplete dropdown for available course codes.
• Chat History:
  • Displays user queries and AI responses.
• Help Section:
  • Toggles a help guide for using the app.

🖥️ Backend

• API Endpoints:
  • /submit-prompt: Accepts user prompts and returns AI-generated responses.
• Textbook Management:
  • Textbooks are uploaded to Gemini for AI use.
  • Automatic re-uploads every 1.5 days using APScheduler.
• AI Model Configuration:
  • Uses Gemini-1.5-Flash for fast and accurate responses.
  • Safety settings ensure content is appropriate.

🧭 Usage

1. Navigate to the Deployed Frontend:
   • Open the Vercel-deployed URL in a browser.
2. Enter a Prompt:
   • Type /course_code followed by your question or statement.
3. Submit:
   • Press the "Submit" button to send the query.
4. View AI Responses:
   • AI-generated answers appear in the chat history.

🛠️ Next Steps

• Chat History Support:
  • Enable users to view and retrieve prior conversations within the application.
• Reducing Model Response Time:
  • Optimize backend processing and API calls for faster AI-generated responses.
• Adding Support for More Courses:
  • Expand textbook integration to cover additional subjects and course materials.
• Multimodal Support:
  • Incorporate image, video, and other formats to enrich learning interactions.