Ducky

Documentation - improve how chunking is explained

Documentation - Improve explanation of the alpha value

Improve documentation when the user is deleting an index

Would be nice to have the dashboard being more responsive, so developers can have windows side by side


UI - Make developer dashboard responsive

Allow ducky to consume PDF files and index them

Add support to index PDF files directly

Sync data directly from another database such as MongoDB or Supabase

Using Ducky with LLM agents as a form of memory with libraries such as LangChain.

Better support for LLM agents

Ducky should be able to recognize JSON documents, without the user having to format them as texts

Add support for indexing JSON documents directly

Add support for images including text extraction

Some indexes may contain multiple identical documents or documents with very similar content. To enhance search accuracy and streamline retrieval results, we'd like to introduce a feature that automatically identifies and deduplicates these documents.

Deduplicate documents in retrieval automatically

Make retrival faster by caching search results, this is useful for multiuser and frequent queries

Retrieval caching

Such as changing team names and invite members

Developer dashboards