> For the complete documentation index, see [llms.txt](https://docs.darcyiq.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.darcyiq.com/organize/projects/interviews.md).

# Interviews

**Interviews** let an Interview Agent run AI-led discovery conversations with people outside DarcyIQ — customers, stakeholders, end users, or prospects. You define what the interview should cover, publish a shareable link, and the agent conducts each conversation over **text or voice**. Every completed session is automatically turned into a structured **discovery document** you can read, edit, and feed back into the project.

{% hint style="success" %}
**Discovery at scale**: Instead of scheduling ten stakeholder calls, publish one interview link. Each participant joins on their own time, the AI Agent walks them through your topics over text or voice, and you get ten structured discovery documents — without booking a single meeting.
{% endhint %}

Interviews live inside a project and appear as the **Interviews** tab on the project detail page.

## Who Is This For?

Interviews are built for any team that needs to gather structured input from many people without manual note-taking:

* **System Integrators & Consultants** — run discovery interviews with client stakeholders before scoping an engagement
* **Product Teams** — collect structured feedback from end users and customers
* **Pre-Sales / Solutions Teams** — capture requirements from multiple stakeholders in parallel
* **Customer Success** — gather onboarding context or quarterly review input
* **Research Teams** — conduct repeatable, consistent interviews at scale

## Key Benefits

| Benefit                      | Description                                                                                        | Business Impact                                               |
| ---------------------------- | -------------------------------------------------------------------------------------------------- | ------------------------------------------------------------- |
| **AI-Led Conversations**     | The Interview Agent asks your questions, follows up naturally, and keeps the conversation on track | Consistent, high-quality interviews without a human moderator |
| **Text or Voice**            | Participants choose to type their answers or just talk (Amazon Nova Sonic voice)                   | Lower friction and higher completion rates                    |
| **Structured Topics**        | Define topics and questions; the agent ticks each one off as it's covered                          | Comparable, structured results across participants            |
| **Self-Serve Link**          | Publish a password-protected public link participants join on their own schedule                   | Eliminate scheduling overhead                                 |
| **Automatic Discovery Docs** | Each completed session produces an editable discovery document                                     | Insights ready to use, not raw transcripts                    |
| **Project Knowledge**        | Optionally arm the agent with the project's knowledge bases and MCP tools                          | Context-aware questions grounded in your data                 |
| **Templates**                | Scaffold a starter set of topics and questions from a template                                     | Launch a well-structured interview in minutes                 |

## How Interviews Work

```mermaid
graph LR
    A["Create<br/>Interview"] --> B["Configure<br/>(3 steps)"]
    B --> C["Publish &<br/>Share Link"]
    C --> D["Participants<br/>Join (text/voice)"]
    D --> E["Sessions &<br/>Discovery Docs"]
```

| Component              | Function                                                      |
| ---------------------- | ------------------------------------------------------------- |
| **Interview**          | A configured discovery conversation that belongs to a project |
| **Interview Agent**    | The project's assigned agent that conducts the conversation   |
| **Welcome Message**    | The greeting participants see before they start               |
| **Topics & Questions** | The structured checklist the agent walks through              |
| **Public Link**        | The password-protected URL participants use to join           |
| **Session**            | One participant's interview run                               |
| **Discovery Document** | The structured artifact generated from a completed session    |

***

## Creating an Interview

{% stepper %}
{% step %}
**Open the Interviews tab** From the project detail page, click the **Interviews** tab.
{% endstep %}

{% step %}
**Click "New Interview"** A new draft interview is created and opens to its configuration page. You can rename it at any time.
{% endstep %}

{% step %}
**Configure the three steps** Work through Interview Details, the User Welcome Message, and Topics & Questions (see below). Changes save automatically.
{% endstep %}

{% step %}
**Publish** Once the welcome message and at least one question are in place, click **Publish** to generate the public link and password.
{% endstep %}
{% endstepper %}

***

## Configuring an Interview

The configuration page is a three-step wizard. You can jump between steps freely, and everything **saves automatically** — there is no manual save button.

### Step 1 — Interview Details

| Field                         | Description                                                                                                                                                                  |
| ----------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| **Interview Name**            | Internal label shown in the project's interview list and the host UI. Not shown to participants.                                                                             |
| **Interview Description**     | Optional internal summary of what the interview is about.                                                                                                                    |
| **Access Password**           | The password participants must enter to join. Auto-generated; copy it with the copy button or edit it.                                                                       |
| **Include Project Knowledge** | When on, the Interview Agent's knowledge bases and MCP tools are attached to the interview chat so it can answer with project context. When off, the agent runs prompt-only. |

### Step 2 — User Welcome Message

A short **Markdown** greeting shown to the participant on the text or voice screen before the conversation starts. Use it to set expectations, explain the purpose, and thank them for their time.

{% hint style="info" %}
The welcome message is **required to publish**. It's the only host-authored copy participants see before they begin, so keep it warm and clear.
{% endhint %}

### Step 3 — Topics & Questions

Define the structured checklist the Interview Agent walks through. Each **topic** groups a set of **questions**, and the agent covers them one by one, asking natural follow-ups as needed.

* **Add topics and questions** manually, reorder them with the up/down controls, and remove any you don't need.
* **Topic description** (optional) gives the agent extra context about what to keep in mind while exploring that area.
* **Templates** — click **Templates** to scaffold a starter set of topics and questions. Some templates include a few quick fields to fill in (for example, a product or company name) that are substituted into the generated questions. Picking a template replaces your current topics, so you'll be asked to confirm if you've already added some.
* **Start blank** — write everything yourself instead of using a template.

At least one topic with at least one question is **required to publish**.

{% hint style="info" %}
The wizard flags steps that still need content with an amber **"Required to publish"** marker, so you can see what's blocking publish at a glance — without having to click Publish first.
{% endhint %}

***

## Publishing & Inviting Participants

### Publishing

Click **Publish** (in the header or from the final wizard step). Publishing generates a public link and confirms the access password. The **Interview published** dialog gives you:

* the **public link** participants follow,
* the **access password**,
* a ready-to-send **invitation message** that bundles both, and
* an **Invite a participant** button that opens the invite dialog.

### Sending Invites by Email

Click **Send Invites** (or **Invite a participant**) to open the invite dialog. Paste one or more email addresses — separated by commas, spaces, or new lines — and each recipient receives a branded email containing the public link and password. The interview must be **published** before you can send invites.

### Status & Lifecycle

| Status        | Meaning                                                                  |
| ------------- | ------------------------------------------------------------------------ |
| **Draft**     | Being configured; not yet joinable. The public link doesn't work yet.    |
| **Published** | Live and accepting participants.                                         |
| **Closed**    | No longer accepting new participants. You can **Republish** at any time. |

Use **Unpublish** in the header to close a live interview. In-progress sessions can still finish, and you can republish later.

***

## The Participant Experience

Participants never see the project, the host UI, or your internal interview name — just a clean, branded interview page.

{% stepper %}
{% step %}
**Open the link** The participant follows the public link.
{% endstep %}

{% step %}
**Enter details** They provide their name, email, and the access password to join. Each participant can join **once**.
{% endstep %}

{% step %}
**Choose a mode** They pick **Text** (type answers in a chat) or **Voice** (talk to the agent out loud). They can switch modes mid-interview, though switching restarts the conversation.
{% endstep %}

{% step %}
**Have the conversation** The Interview Agent greets them with your welcome message and works through your topics. A progress panel shows which topics have been covered as the conversation unfolds.
{% endstep %}

{% step %}
**Finish** The agent wraps up when it has covered everything, or the participant can end early. They see a simple "Thank you!" confirmation.
{% endstep %}
{% endstepper %}

{% hint style="info" %}
Voice responses are transcribed to text, so results are captured consistently regardless of which mode the participant chooses. Suggest participants set aside \~15–30 minutes of uninterrupted time.
{% endhint %}

***

## Sessions & Discovery Documents

The **Sessions** tab on the interview detail page shows every participant who has joined.

### Tracking Sessions

* **Search** participants by name or email.
* **Status** is shown as a pill: **Not started**, **In progress**, or **Completed**, along with start/completion times and the session duration.

### Discovery Documents

When a session completes, the interview generates a **discovery document** — a structured artifact summarizing the participant's answers, organized around your topics and questions. From the Sessions tab you can open the document in a side panel to **read, edit, and refine** it. Because discovery documents are standard artifacts, they can be reused anywhere artifacts are used across the project.

***

## Best Practices

| Practice                        | Recommendation                                                                                            |
| ------------------------------- | --------------------------------------------------------------------------------------------------------- |
| **Keep topics focused**         | A handful of clear topics with 2–4 questions each beats one long, sprawling checklist.                    |
| **Write a welcoming greeting**  | The welcome message sets the tone — explain the purpose and how long it will take.                        |
| **Start from a template**       | Templates give you a well-structured baseline you can refine, instead of a blank page.                    |
| **Enable Project Knowledge**    | Turn it on when participants may ask about your product or project so the agent can respond with context. |
| **Test the link yourself**      | Open the public link in a new tab and run through the join flow before sending invites.                   |
| **Review discovery docs early** | Read the first few completed sessions to confirm the topics are eliciting the answers you need.           |
| **Close when you're done**      | Unpublish the interview once you've collected enough responses; republish later if you reopen it.         |

{% hint style="success" %}
**Pro Tip**: Pair Interviews with the rest of the project. Run discovery interviews during the sales or scoping phase, then use the generated discovery documents as context when you generate the project [Definition & Specs](/organize/projects/definition-and-specs.md) — your requirements are grounded in what participants actually told you.
{% endhint %}

## Next Steps

| Goal                                       | Documentation                                                    |
| ------------------------------------------ | ---------------------------------------------------------------- |
| Set up the project that hosts interviews   | [Projects](/organize/projects.md)                                |
| Turn discovery into requirements           | [Definition & Specs](/organize/projects/definition-and-specs.md) |
| Learn about the agents that run interviews | [Agents](/core-features/agents.md)                               |
| Configure knowledge bases for context      | [Knowledge Bases](/additional-features/knowledge-bases.md)       |


---

# 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, and the optional `goal` query parameter:

```
GET https://docs.darcyiq.com/organize/projects/interviews.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

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.