Skip to main content
Replyke’s Search & AI features use vector embeddings to power natural language search and question answering across all content in your project — entities, comments, chat messages, users, and spaces.
Semantic search and AI features require embeddings to be enabled for your project. Configure this in the Replyke dashboard under Settings → Semantic Search / AI.
Search returns the most semantically similar records to a given query. Results include a similarity score alongside the matched record.

​
AI Q&A

useAskContent lets users ask natural language questions and receive synthesized answers generated from your project’s content. The response streams token by token via SSE (Server-Sent Events) and includes source citations.

​
In This Section

Semantic Search

Search entities, comments, messages, users, and spaces by natural language query

Ask

Ask a question and stream a generated answer with source links

​
Hooks

useSearchContent

Search entities, comments, and messages by natural language

useSearchUsers

Find users by name, bio, or other profile text

useSearchSpaces

Find spaces by name, description, or topic

useAskContent

Ask a question and receive a streamed, cited answer