Frequently Asked Questions

What is Petrichor?

Petrichor is an advanced writing and research application designed to transform traditional note-taking into a dynamic, interconnected knowledge network. It automatically links notes using semantic analysis, offers an intuitive WYSIWYG editor with full Markdown and LaTeX support, and integrates an AI assistant built on a Retrieval Augmented Generation (RAG) framework to provide context-aware insights.

How does automatic linking work?

Petrichor uses cosine similarity algorithms to assess semantic relationships between notes. This method enables automatic linking, where the system identifies and connects related content without manual intervention. Additionally, when links to external resources are added, their contents are scraped and indexed in a vector database, ensuring that both internal notes and external information are cohesively integrated.

How does the interactive graph view function?

The interactive graph view is a core feature of Petrichor that visually represents your note ecosystem. Key aspects include:

• Dynamic Node Representation: Each note or external resource is visualized as a node.
• Interactive Summaries: Hovering over a node reveals a concise summary of the associated content.
• Seamless Navigation: Clicking on a node opens the full note in the editor, allowing for fluid exploration of related ideas.
• Integrated Resources: Both user-generated content and scraped external resources are part of the network, providing a comprehensive view of your research.

How does the AI assistant operate?

The AI assistant in Petrichor is powered by an advanced RAG framework. It works by:

• Contextual Data Retrieval: Accessing and processing the interconnected network of notes and external resources.
• Advanced Language Modeling: Utilizing state-of-the-art language models to generate precise, context-aware responses.
• Seamless Integration: Automatically linking AI-generated insights back to the relevant notes, ensuring that every piece of information is interconnected within your research.

What technical stack is used in Petrichor?

Petrichor is built using modern, high-performance technologies:

• Frontend: A responsive web interface that supports real-time rendering of Markdown, LaTeX, and Mermaid diagrams.
• Backend: Core services are engineered in Rust for efficient processing and robust data handling.
• AI & Machine Learning: Integration of state-of-the-art language models and vector databases to support the RAG framework.
• Data Security: Secure data storage and encrypted communications are standard, ensuring that your notes and research data are protected.

Will Petrichor be open-sourced?

While there is a possibility of open sourcing Petrichor in the future, the current version remains closed source. For now, all features are available free of charge, and the project continues to evolve with user feedback and new developments.

What platforms are currently supported?

At present, Petrichor is available as a web-based application. I plan on adding dedicated desktop and mobile applications to extend support across multiple devices, ensuring a seamless and consistent user experience regardless of your preferred platform.

Is there any cost associated with using Petrichor?

Everything in Petrichor is free at the moment. There are no subscription fees or hidden costs, and all functionalities are available to users at no charge. Future pricing models have not been explored yet, as our focus remains on enhancing and expanding the platform.

How is data security handled in Petrichor?

Data security is a top priority:

• Secure Storage: Your notes and research data are stored using industry-standard encryption protocols.
• Encrypted Communications: All data transmissions are securely encrypted.
• Best Practices: We adhere to rigorous security practices to protect your intellectual property and personal data.

What does the future roadmap look like?

Our upcoming major feature is Deep Research Mode, which will:

• Enhance Web Integration: Empower the AI assistant with the ability to search and scrape the web for relevant information.
• Automate Detailed Reports: Automatically generate comprehensive research notes that link back to the original queries.
• Expand Platform Support: Continue to roll out performance improvements and additional platform-specific features, including native desktop and mobile apps.
• LLM Tools: Expand the AI assistant's capabilities and expand upon the current RAG framework.

How can I provide feedback or report issues?

We value your input and strive to continuously improve Petrichor. Please feel free to reach out to us via email at [email protected] with any feedback, suggestions, or bug reports.