Summary
perplexitree transforms learning into an engaging visual experience by combining puzzle mechanics with AI-powered knowledge exploration. Players cultivate knowledge trees through strategic pruning while leveraging Perplexity's real-time search capabilities.
Technical Approach: Built with FastAPI backend and vanilla JavaScript frontend, the application uses a two-phase Perplexity integration: initial queries use the sonar-pro chat completion model for structured knowledge area generation, while subsequent growth uses the search model with negative prompting to ensure unique web results.
Impact: The negative prompting system prevents information redundancy by maintaining a history of explored subtopics and excluding them from future queries. This allows users to dive deep into specific branches for detailed research while simultaneously branching out to explore adjacent topics. Players also have the freedom to prune unwanted knowledge branches, curating their own learning path. By combining Perplexity's search capabilities with context-aware filtering, the system addresses information overload through an intuitive tree metaphor, presenting diverse insights that build comprehensive understanding while giving users full control over their exploration.
Built With
- fastapi
- html5
- javascript
- perplexity-ai
- python
- sqlite
- sqlite-database
- vercel
Log in or sign up for Devpost to join the conversation.