# Underheard Research

> Underheard Research runs AI-moderated interviews (spoken or written) with a company's own people or with the Underheard panel, and turns them into analysed results and reports. This file is the index of its manual; every page is also served as plain Markdown.

Start with the overview, then the page for what you are doing. The REST API and the MCP server take the same key, made in Settings under API. The manual is public and holds no private study data.

## Getting started

- [Research documentation](https://research.underheard.ai/llms/overview.md): Plan a study, hear from the right people, and turn their answers into findings you can use.
- [Your first study](https://research.underheard.ai/llms/quick-start.md): A practical path from the question you need to answer to a report you can share.
- [Study design](https://research.underheard.ai/llms/study-design.md): Build a clear guide, capture measured answers, and check the journey before fieldwork.

## Running research

- [Audiences and quotas](https://research.underheard.ai/llms/audiences.md): Define who belongs in the study and keep the target separate from the available sample.
- [Recruitment and links](https://research.underheard.ai/llms/recruitment.md): Choose who a study recruits: through Underheard, or your people. The links that bring each person in.
- [The respondent experience](https://research.underheard.ai/llms/respondents.md): Understand what people see, what they agree to, and how to check the journey.
- [Fieldwork and quality](https://research.underheard.ai/llms/fieldwork.md): Follow collection, distinguish completed from accepted responses, and investigate problems.

## Understanding results

- [Analysis and evidence](https://research.underheard.ai/llms/analysis.md): Explore findings while keeping the sample, source answers and limits in view.
- [Quantitative methods](https://research.underheard.ai/llms/methods.md): Choose the right calculation and make its assumptions visible in the deliverable.
- [Reports and exports](https://research.underheard.ai/llms/reports.md): Prepare a readable report and deliver files that preserve the findings and their scope.
- [Library and past research](https://research.underheard.ai/llms/library.md): Reuse what you have learned while preserving the context of each study.

## For companies

- [Working with a research partner](https://research.underheard.ai/llms/working-with-agencies.md): A shared checklist for commissioning, reviewing and delivering a study.
- [Privacy and consent](https://research.underheard.ai/llms/privacy.md): Identify the participation, recording and data-handling choices that a study needs.
- [Troubleshooting](https://research.underheard.ai/llms/troubleshooting.md): Find the failing stage, preserve your work, and collect useful information for support.

## Integrations

- [Integration planning](https://research.underheard.ai/llms/integrations.md): Choose a workflow, make a key, and know which door it opens.
- [API access](https://research.underheard.ai/llms/api.md): Which door opens what: the REST API for people and links, the MCP server for everything else, and what stays on screen.
- [API keys](https://research.underheard.ai/llms/api-keys.md): Make a key in Settings. One key opens the REST API and the MCP server, as you, in your workspace.
- [REST API](https://research.underheard.ai/llms/rest-api.md): Add people to a study, prepare and download personal links, and read a study's links from your own systems.
- [Assistants and MCP](https://research.underheard.ai/llms/mcp.md): Connect Claude, ChatGPT, Cursor or any MCP client to your workspace, with the full tool catalogue, generated from the server.
- [For agents](https://research.underheard.ai/llms/agents.md): How an AI agent should read this manual and work with the platform, with recipes for the common jobs.

## Everything in one file

- [Full manual](https://research.underheard.ai/llms-full.txt): all pages above, in reading order.