> ## Documentation Index
> Fetch the complete documentation index at: https://docs.dialog.tools/llms.txt
> Use this file to discover all available pages before exploring further.

# Research Chat

> The core research experience - ask questions, get sourced insights

The research chat is Dialog's primary interface. Ask natural language questions and receive structured, source-backed research briefs in real time.

## How it works

When you submit a query, Dialog:

1. **Analyzes your question** to determine the best research approach
2. **Discovers relevant communities** across Reddit and other sources
3. **Fetches and analyzes posts and comments** from those communities
4. **Synthesizes findings** into a structured brief with citations

All of this happens in real time - you can watch the research unfold as it streams back to you.

## Real-time streaming

Dialog uses server-sent events to stream results as they're generated. You'll see:

* **Tool activity** - Which data sources are being queried
* **Thinking process** - The AI's reasoning as it analyzes data
* **Final output** - The structured research brief with citations

## Session history

Every research session is automatically saved to your account:

* **Persistent across devices** - Sign in anywhere to access your history
* **Searchable** - Browse past sessions in the sidebar
* **Contextual** - Follow-up questions maintain the full conversation context

## Activity tracking

Each query shows activity indicators so you can see exactly what Dialog did:

* Which MCP tools were invoked
* How many sources were consulted
* The execution timeline

## Tips for best results

<Tip>
  Be specific about what you want to learn. "What do people think about X?" is good. "What are the top 3 complaints about X's pricing on Reddit?" is better.
</Tip>

<Tip>
  Use follow-up questions to drill deeper. Dialog maintains full context within a session, so you can refine your research iteratively.
</Tip>

<Warning>
  Very broad queries (e.g., "Tell me everything about AI") may take longer to process and produce less focused results. Narrow your scope for better output.
</Warning>