AiinakDocs

Command Palette

Search for a command to run...

Conversational Search

Chat sessions over WebSocket, search history, autocomplete suggestions and search analytics.

Chat sessions

Every search can become a conversation. AI Search holds a real-time chat session over WebSocket, so follow-up questions stream back instantly and carry the full context of what came before — "what about its competitors?" works because the session remembers which company you were discussing.

Session management

Conversations are saved as sessions you can name, revisit and continue later from where you left off.

Real-time streaming

The WebSocket connection streams each answer as it is generated, including progressive enhancements.

Context retention

Entities, sources and answers from earlier in the session inform every follow-up.

History and suggestions

Your search history is kept and searchable, so you can re-open any past query and its answer. History also powers autocomplete: as you type, AI Search suggests completions based on what you have searched before, getting more useful the more you use it.

  1. 1

    Open History

    Browse past queries and chat sessions chronologically, or filter by keyword.

  2. 2

    Re-open a session

    Continue the conversation with full context, or branch into a new direction.

  3. 3

    Prune when needed

    Delete individual entries or clear history entirely — removed entries stop feeding suggestions.

Search analytics

The analytics dashboard tracks your queries, performance metrics and user intent classification — what you search for, how fast answers arrive, and which kinds of questions (informational, navigational, transactional) dominate your usage.

MetricWhat it tells you
Query trackingVolume and recurring topics across your searches
Performance metricsAnswer latency and how often progressive enhancement upgraded results
Intent classificationThe mix of question types you ask, useful for tuning team workflows