EduVerse

Imagine an education platform uniting students, parents, and teachers with AI superpowers. Presenting EduVerse: Where learning evolves, together.

Comment

Inspiration

Imagine a world where every child, regardless of their background, has access to a quality education that ignites their passion and empowers them to reach their full potential. No child left behind. EduVerse bridges the gap, offering:

DIGITAL EDUCATION CAN TRANSFORM THE LIVES OF ~20 LACS STUDENTS IN INDIA!!

  1. Multilingual learning: Knowledge for all, regardless of language.
  2. Personalized AI: Tailored paths for every student to shine.
  3. Teacher support: AI empowers educators, frees them to inspire.
  4. Mental health focus: Building well-rounded individuals, not just minds.

Let EduVerse be the spark that ignites a revolution in education, one where every child has the opportunity to dream big, reach their full potential, and contribute meaningfully to a brighter future. Join us in this fight for social justice, one student, one teacher, one parent at a time.

Remember, education is not a privilege, it's a right. Let's make it a reality for all.

What EduVerse Does?

Here's a user-flow for each user role describing all the features: eduverse_userflowdiagram

  1. Connects: Facilitates seamless communication and collaboration between students, teachers, and parents through an accessible multilingual platform.
  2. Personalizes: Leverages AI (GenAI) to provide each user with tailored experiences based on their needs, progress, and preferences.
  3. Engages: Offers innovative features like drowsiness alerts, sleep-learning AI, animated sign language videos, and conversational bots to make learning interactive and enjoyable.
  4. Optimizes: Supports teachers with grading assistance, text-to-slide generation, and personalized tips. Provides parents with mental health reports and easy connection options.
  5. Empowers: Equips students with tools like math chatbots, news updates, and reading assistance bots to become independent learners.

EduVerse aims to transform the educational experience by fostering a connected, personalized, and technology-driven ecosystem that benefits all stakeholders.

How we built it?

The overall system architecture of the application is as follows:

eduverse_system_architecture

The tools and technologies are as follows:

  1. Frontend: Reactjs, CSS (AntDesign)
  2. Backend & Database: Nodejs, Python & Firebase
  3. APIs: GoogleTranslate, GoogleClassroom, OpenAI, News, WolframAlpha, GoogleReadAlong, FreeDictionary, GoogleTTS, Play.ht, FastAPI, GoogleCloud Vision
  4. 3rd Party Tools: Dialogflow, Manim, Marp, StreamLit
  5. Backend Deployment: Render, Google CloudRun
  6. Frontend Deployment: Vercel

The architecture diagrams of other major AI features are as follows:

  1. SignimateAI: If I’m a student who finds traditional lectures boring and seeks visual learning ways to grasp the information, we have, SignimateAI: Give any text prompt and it generates an animated video with sign language generation. Behind this goes Manim code generation using LLM models converting it into animated video via Google Cloud TTS and mapping each letter wrt sign language.

eduverse_system_architecture

  1. MorpheusAI: One of the unique features of our application, Morpheus AI. Many students struggle with remembering information. According to research papers, in the REM phase, memory retention is maximum with auditory stimuli, and we have done just this. With any text prompt, a story is generated using GenAI which is converted to a peaceful and calm speech using play.ht api and this generated audio is resampled with sleeping waves audio to generate the final audio.

eduverse_system_architecture

  1. Text prompt to slides generation:

eduverse_system_architecture

  1. AI Grading Assistant:

eduverse_system_architecture

Challenges we ran into

  1. Limited computational resources to run the models and servers faster.
  2. Restrictive API limits in the free tier of the APIs.
  3. Dialogflow ES web integration has a 5-second timeout limit.

Accomplishments that we're proud of

Technical implementation of all the features we ideated and DEPLOYING the entire application (frontend, backend & ML models)!!!

Our attempt to make the data generated by LLMs more reliable: we did it by few-shot prompting and community feedback. Here's the methodology.

What we learned

  1. Utilizing free tiers can limit functionality.
  2. We encountered limitations with Dialogflow ES web integration.
  3. Exploring solutions like cloud-based scaling and efficient API utilization is crucial. Some of our backend servers deployed on the free tier of Render led to some features of the application being very slow, thus, forcing us to shift to deploying them on Google Cloud Run. (having 200$ free credits of Google Cloud was a savior)

What's next for EduVerse

  1. Multilingual support for Chatbots
  2. Making the application scalable and connecting to high computational databases like AWS.

Built With

Share this project:

Updates