• 

  index

• 

  Dashboard

Omaricha

Inspiration

The idea for Omaricha was born out of a challenge many students face: how to efficiently manage heavy academic workloads while truly understanding the material. Traditional learning tools often lack adaptability and interactivity. We wanted to change that by building an intelligent, conversational study assistant that supports students in real time. With the power of Gemini API and open-source embedding models, Omaricha transforms how students engage with content—making learning smarter, faster, and more personalized.

What it does

Omaricha is an AI-powered academic assistant that helps students study more effectively. It allows users to:

• Upload documents (PDFs, notes, slides) and automatically generate dynamic quizzes from them.
• Interact via voice, asking questions and receiving instant, conversational answers.
• Listen to documents read aloud by AI—ideal for multitasking or auditory learners.
• Receive real-time grading and feedback on quizzes.
• Work with various quiz formats: multiple choice, fill-in-the-blank, true/false, and inline questions.

How we built it

• Gemini API powers Omaricha’s conversational intelligence and summarization capabilities.
• Open-source embedding models (e.g., SentenceTransformers, OpenAI-compatible encoders) are used to semantically analyze and extract key information from user-uploaded documents.
• We built a custom quiz generation engine that creates questions based on topic relevance and difficulty.
• Voice interaction is enabled through speech-to-text and text-to-speech APIs for seamless engagement.

Challenges we ran into

• Extracting meaningful context from a wide variety of document types.
• Ensuring quiz accuracy and relevance from unstructured inputs.
• Balancing speed with deep semantic analysis for real-time interactions.

Accomplishments that we're proud of

• Building a full pipeline from document upload → embedding → quiz generation → real-time interaction.
• Creating a natural voice interface using Gemini and speech APIs.
• Designing a system that adapts to any course or subject—completely user-driven.

What we learned

• How to integrate Gemini API with open-source tools for hybrid AI systems.
• The power of embeddings in creating dynamic learning experiences.
• The importance of UX in educational tools—AI is only as good as how intuitively it's delivered.

What's next for Omaricha

• Launching mobile and browser extensions for on-the-go learning.
• Adding support for more file types and languages.
• Improving personalization through adaptive learning paths and progress analytics.

Built With

• gemini
• postgresql
• superbase