The following is a list of features currently supported.
The following frameworks are currently supported.
Your documentation is ingested at build time, transformed, embedded, and indexed for search.
- Ingesting at build time ensures your search index is always up-to-date with the latest documentation changes.
- Built on Nextra's existing search integration and Docusaurus' plugin system, documentation is then delimited by section headings to provide more nuanced results and include as context for the AI-assistant.
- A vector representation is generated and persisted to a vector store search index for each section.
- The search index powers both the documentation search and the AI-assistant.
Our UI components are rendered in place of your documentations default search component, we take care of the rest.
These components are built with battle tested primitives to ensure accessibility, responsiveness, and performance.
Dark-mode and light-mode are supported out of the box.
Your documentation can be searched with natural language queries and results are ranked by similarity.
Results are grouped and sorted first by the page, then by the section, to provide a structured search experience when exploring at a high-level.
Your documentation readers can ask arbitrary, specific questions and receive direct answers based on your existing documentation using a large language model to synthesize the relevant documentation and provide a direct answers getting them their answer faster.
The following are features that we are considering. If you have any suggestions or feedback, please reach out!
- Framework-specific integrations: search UI components and data ingestion plugins.
- Documentation scraper: support any static documentation by regularly scraping the site and updating the index instead of a build time integration. This would eliminate the need to integrate a plugin and upload the documentation at build time.
- Versioning: support versioning so multiple versions of the same documentation can be live at once, and maintain data consistency when deploying.
- Query analysis: analyze what users are searching for to improve the documentation, identify gaps, and generate suggestions.
- Component styling: support theming or styling to match your documentation's branding.