> For the complete documentation index, see [llms.txt](https://planck-ai.gitbook.io/planck-ai-docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://planck-ai.gitbook.io/planck-ai-docs/documentation/core-concepts/conversations.md). # Conversations A conversation is a chat thread inside a workspace. Conversations let users ask questions, review answers, and continue with follow-up questions. The conversation keeps context from earlier messages, so users do not need to repeat everything in each question. ### What conversations are used for Use conversations when you want to: * Ask questions about documents * Query connected databases * Analyze spreadsheets or CSV files * Compare information across sources * Generate summaries, tables, charts, or reports * Continue an analysis over multiple follow-up questions ### Stateless chat vs conversations Planck AI supports two common chat patterns. #### Stateless chat A stateless chat request does not require a saved conversation. Use this when you want a simple one-off answer. Example: ``` Summarize this document. ``` #### Stateful conversations A stateful conversation saves the chat history. Use this when the user may ask follow-up questions. Example: ``` How has revenue changed in the last two quarters? ``` Follow-up: ``` Break that down by product category. ``` Another follow-up: ``` Show the result as a chart. ``` ### Asking good questions Good questions are specific and include the business context where needed. Good examples: ``` Summarize the main risks in this contract. ``` ``` Which suppliers had the highest spend this month? ``` ``` Compare planned rollout milestones with actual completion dates. ``` ``` Show delayed projects by region. ``` Less useful examples: ``` Analyze this. ``` ``` What is going on? ``` ``` Tell me everything. ``` ### Types of answers Depending on the question and available data, Planck AI may return: * Text explanations * Tables * Charts * SQL or code snippets * Maps * Document references * Source context ### Follow-up questions Follow-up questions are useful when you want to refine or extend an answer. Examples: ``` Show this as a table. ``` ``` Only include records from this month. ``` ``` Break this down by region. ``` ``` Which source supports this answer? ``` ``` Export this as a summary I can send to my team. ``` ### Source context For document and data workflows, Planck AI may include supporting context or source references. Users should review the source context before relying on an answer for business-critical decisions. ### Best practices Use one conversation for one topic or workflow. For example, create separate conversations for: * Monthly finance review * Contract analysis * Network rollout delays * Supplier spend analysis * Customer support reporting Keeping conversations focused makes it easier to follow the analysis and reuse the thread later. --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter: ``` GET https://planck-ai.gitbook.io/planck-ai-docs/documentation/core-concepts/conversations.md?ask= ``` The question should be specific, self-contained, and written in natural language. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.