-
-
User login page
-
User signup page
-
An interactive domain selector offering 4 categories—Fun, Finance, Health, and Research—for a personalized AI experience
-
Seamless Transitions to chatpage
-
Chat Page
-
Dark Theme
-
Persona by age
-
persona selection
-
History and chat grouping (Dark Theme)
-
Project grouping and history search
-
Library for showing downloaded tools and techniques(i.e. flashcard,roadmap,etc,.)
About the Project: Infinenix
Inspiration
Modern AI tools answer questions — but what if they could understand your purpose and generate an entire learning experience tailored to your intent?
We were inspired by the idea that not every question has a single, definitive answer. Someone asking “What is quantum entanglement?” may want to learn, teach, compare, or dive deeper. We envisioned a system that doesn’t just respond, but interprets, enhances, and visualizes — automatically.
Thus, INFINENIX was born: an AI-powered React web app that understands user intent and generates an immersive visual output using the Perplexity Sonar API.
What It Does
Infinenix is a dynamic, intelligent chat interface that:
- Identifies the user’s intent from natural language prompts (e.g., learn, explore, compare, summarize).
- Matches the intent to the most suitable enhancement technique (like flashcards, roadmaps, or timelines).
- Uses Sonar API to:
- Generate a custom explanation tailored to the detected intent.
- Generate static, visually-rich Tool and Technique (like flashcards, diagrams, guides).
- Renders the result in the browser, allowing users to immediately preview and download a complete, standalone visual learning experience.
How We Built It
- Frontend: Built using React and Tailwind CSS for a clean, fast, and responsive UI.
- Sonar Integration:
- User prompts are analyzed and reformulated internally into optimized questions and instructions.
- These are sent to the Perplexity Sonar API to fetch both:
- A clear, intent-aware explanation.
- A matching static visual tools and technique response (e.g., interactive flashcards or study roadmaps).
- Intent Detection & Prompt Optimization: Though the backend is abstracted, the frontend leverages smart heuristics to guide Sonar with the right context.
Challenges We Ran Into
- Designing prompts that could consistently elicit both an explanation and structured HTML/CSS output.
- Calibrating Sonar's flexibility: balancing creativity with consistency in how it outputs code + reasoning.
- Ensuring HTML/CSS returned by the API was clean, visually sound, and renderable without errors.
- Integrating the flow in a single-page app so users could get results and visualize them instantly without reloading or configuration.
Accomplishments That We're Proud Of
- Created a fully functioning web app that builds and renders AI-generated content in real-time.
- Transformed the Sonar API into a tool for not just question answering — but intent-based learning enhancement.
- Enabled users to visually experience answers in the form of interactive learning elements, all powered by AI.
- Achieved a seamless chat-to-visual-output workflow using only the frontend and API calls.
What We Learned
- Prompt design is critical when using reasoning APIs like Sonar — the right format unlocks multi-purpose outputs.
- Sonar can act as both a semantic reasoner and a web content generator — enabling experiences far beyond static Q&A.
- Even without a backend, powerful tools like Sonar enable frontend-only apps to perform complex research-style transformations.
- There's massive potential in intent-driven UX where the system adapts how it responds depending on user goals.
What’s Next for Infinenix
- Expand intent classification using lightweight ML models or fine-tuned transformers.
- Add multi-turn chat and deeper reasoning using Sonar Deep Research.
- Support more output types — e.g., infographics, checklists, decision trees.
- Allow users to connect multiple visual outputs into personalized knowledge journeys.
- Launch a mode for educators and researchers to generate self-contained learning modules from questions.
Infinenix reimagines the future of AI interaction: where asking a question builds an experience. Powered by Perplexity’s Sonar API, it’s an intent-aware, visually-enhanced, and learning-first AI assistant — all in one lightweight web app.
Built With
- api
- eslint
- fastapi
- flask
- flask-cors
- git
- github
- javascript
- machine-learning
- netlify
- nltk
- npm
- numpy
- pandas
- pip
- pydantic
- python
- python-dotenv
- pytorch
- react
- requests
- scikit-learn
- sentence-transformers
- shadcn/ui
- spacy
- sumy
- tailwind
- three.js
- torch
- tqdm
- typescript
- uvicorn
- vercel
- vite
- yake
Log in or sign up for Devpost to join the conversation.