Conversational Q&A

Conversational Q&A provides a dedicated interface for extended, multi-turn research conversations. Unlike the sidebar Research AI Panel, which is designed for quick questions alongside other tasks, Conversational Q&A is a full-page experience optimized for deep exploration of research topics.

How It Works

Navigate to the Conversational Q&A page and start a new conversation. The system maintains full context across turns, so you can ask a series of related questions that build on each other. Each response includes citation tracking, and the AI identifies gaps in the available literature as it works.

Free-Tier Limit

Hosted Free accounts currently get 5 Conversational Q&A requests per week.

Depth Control

Conversational Q&A offers three depth levels that control how detailed the AI's responses are:

Brief -- Concise answers of one to three sentences. Useful for fact-checking or quick lookups. The AI provides the core answer without extensive elaboration.

Standard -- Balanced responses with supporting context and citations. This is the default level and works well for most research questions. The AI explains its reasoning and points to relevant sources.

Comprehensive -- Detailed, thorough responses that explore multiple perspectives, methodological considerations, and related findings. Use this level when you are writing a literature review or need to understand a topic in depth. Responses at this level may be significantly longer and take more time to generate.

Select your preferred depth level before sending a question. You can change the level between turns to get a brief overview first and then request a comprehensive deep-dive on a specific aspect.

Citation Tracking

Every factual claim in a Conversational Q&A response is linked to a source. The system tracks citations across the entire conversation, building a cumulative reference list as the dialogue progresses. This is particularly valuable for literature reviews, where you need to maintain a clear record of which papers support which claims.

Citations include:

• Document title and author information
• Section or page references
• A confidence indicator showing how strongly the source supports the claim

Gap Identification

As the AI searches your document library for answers, it identifies areas where the available literature does not fully address your question. These gaps appear as explicit callouts in the response, helping you:

• Recognize limitations in your current reading list
• Identify areas where additional research is needed
• Frame future research questions around unresolved issues
• Spot opportunities for original contributions

Gap identification is automatic. The AI does not need a specific prompt to flag missing coverage -- it does so whenever it detects that your question touches on areas not well-represented in your documents.

Multi-Turn Conversations

Conversational Q&A is designed for extended dialogues. The system maintains context across turns, so you can:

• Ask follow-up questions that reference earlier answers
• Refine your focus progressively, starting broad and narrowing down
• Request comparisons between points raised in different parts of the conversation
• Build a structured understanding of a topic through iterative questioning

The AI tracks the full conversation thread and adjusts its responses based on what has already been discussed.

Tips

• Use depth control strategically. Start with Brief for orientation, switch to Standard for working questions, and use Comprehensive for the most important topics in your research.
• Pay attention to gap identification callouts. They often point to the most productive areas for future reading or original research.
• Conversational Q&A works best when your document library is well-populated. Upload relevant papers before starting a deep conversation to give the AI more material to draw from.
• Export the citation list from a Conversational Q&A session to seed your reference manager with sources identified during the conversation.