> ## Documentation Index > Fetch the complete documentation index at: https://conveo.ai/docs/llms.txt > Use this file to discover all available pages before exploring further. # Study Personas > Generate participant profiles, check their support, and explore simulated perspectives ## What study personas are Study personas summarize patterns in your interview participants as fictional profiles. A generated profile includes a name, description, characteristics, and available supporting quotes or classified interviews. Use personas to discuss audience differences while keeping the underlying research accessible. The study **Personas** page generates personas from study data. [Organization personas](/docs/insights-and-reports/persona-library) are a separate capability for reusable profiles. Creating a profile there is not the same as classifying this study into a predefined segmentation. ## Generate personas When Personas is enabled for your organization, open **Personas** in the study navigation and select **Generate personas**. Study analysis must be available. Generation runs in the background and the page shows progress or a failure state. The study needs enough distinct participants with completed, visible, non-test interviews to meet your organization's minimum. Multiple diary entries from one participant do not count as multiple participants. The page displays the effective minimum instead of a universal fixed interview count. ## Review participant support Review each persona's description, quotes, and associated interviews. Check how many distinct participants support the profile and how that compares with the study total. Personas below the organization's minimum support are not surfaced. If none qualify, the page explains that no personas met the threshold. Collect more participants, or ask an organization admin to review **minimum participants per persona** in Organization settings before regenerating. A separate low-support warning marks personas that are still shown but should be treated as directional findings. A profile's large percentage within a small study does not make the sample robust. ## Explore the source evidence Use the persona's interview list and quoted evidence to check whether its traits fit the participants. Names and profile images represent the synthesized persona; they are not identification of a real participant. Treat the classification as an AI interpretation of this study's sample. It is not a statistically weighted estimate of population segments. ## Chat with a persona Select **Chat** beside a persona. Chat requires signing in and permission to use persona chat. Ask questions to explore the perspective described by the profile. Persona chat generates responses from the persona's description, characteristics, and conversation context. It does not retrieve the original transcripts to substantiate each answer. A persona's response to a new idea or image is a simulation, not a response collected from participants. Do not present generated dialogue as a participant quote or use it as proof that customers will behave a certain way. ## Changes and limitations Personas depend on the quality and coverage of the source interviews. Missing evidence cannot be repaired by a more detailed fictional profile. Regenerating personas can change their composition. Replacing a generated set removes the previous study personas and their associated chats. Save any material you need and clearly label simulated dialogue before a regeneration. Verify research claims in the original interviews or use [Talk to your data](/docs/interview-data/talk-to-your-data) to investigate evidence. Use [Facets](/docs/insights-and-reports/facets) for explicit categories you need to count and compare.