# About Human Virtual

> Your knowledge, presented live, while you're somewhere else.

App: https://app.hv.ai

Human Virtual is a studio for building an AI Avatar that presents for you. You give it a face, a voice and what you know. It turns that into a presentation, then shares it as a link. Everyone who opens the link gets their own live conversation with your avatar: it presents, they interrupt, it answers. Afterwards you see who came, what they asked and how it landed.

## How the pieces fit

### Avatar

**The presenter. The one who speaks for you.**

An avatar has a name, a role, a look and a voice, and its own purpose and knowledge. It is the same presenter in every flow and every room you put it in. When it presents, it speaks the flow's script; when a visitor asks something, it answers as itself.

- Use a photo of yourself, with your consent, or choose a ready-made presenter or animated character.
- Give it standing instructions: how to speak, what to stress, what to avoid. It follows them in every room without reading them out.
- Give it a purpose, knowledge and training: what it is there to do, the facts, files and web pages it knows, and Q&A pairs, including answers to what visitors asked it. These shape its live answers in every room, not the script it narrates.
- It is always introduced as an AI Avatar. It represents you and never pretends to be you.

### Voice

**How your avatar sounds, in every language it speaks.**

Record a short sample and Human Virtual clones your voice, or pick one from the shared voice library. One voice can speak all nine supported languages.

- Every avatar has a default voice, used for any language without a voice of its own.
- Add a voice for a specific language and it replaces the default for that language, in narration and in live answers.
- Supported languages: English, Vietnamese, Spanish, French, German, Japanese, Korean, Chinese and Portuguese.

### Flow

**The presentation itself: an ordered set of scenes.**

Describe what you want, attach your deck, documents or links, and the AI builds a flow you can edit. Each scene is a slide-like screen with what your avatar says over it.

- Edit every scene in the studio: its text, its layout, its narration and which avatars appear on it.
- Add knowledge to the whole flow or to one scene; the avatar brings its own knowledge too. It answers only from approved knowledge and says when a question falls outside it.
- Export any flow as a video, or publish it as a Live Room.

### Live Room

**A link where your avatar presents a flow to anyone who opens it.**

Each visitor gets their own conversation in real time. The avatar plays the scenes, listens, and answers by voice. "Live" means it responds live, not that you have to be there.

- Pin the room's language, or let the avatar reply in each visitor's language.
- Lock a room with an access code, or address it to one person so the avatar greets them by name.
- Scenes can move on by themselves or wait for the visitor.
- Each room keeps a report: visits, conversations, time spent, what people asked most, and how it landed.

## Step by step

The everyday tasks, in the order most creators meet them. Names in bold are the buttons and sections you see in the app.

### Create an avatar

1. On the dashboard, under **Avatars**, click **Create avatar**. In the editor, **Create new avatar** in the Avatar panel opens the same four steps.
2. **Name & Role**: name the avatar, give it a role and choose its **Primary language**, the language of the sentence you read for the voice sample.
3. **Photo**: **Use Camera** or **Upload File**, then drag and zoom until the face fills the square. **Remove background** is optional.
4. **Voice**: **Record my voice** and read the sentence aloud, or **Pick a voice** from the library.
5. **Review**, then **Create Avatar**. A recorded voice is cloned automatically after that.

### Create a flow

1. On the dashboard, under **Flows**, click **Create with AI**.
2. Describe what your avatar should present, and to whom. **Add sources** attaches documents, images or a link.
3. Choose the avatar, or keep **Best fit**, and the stage shape. The quote shows the most credits the build can cost.
4. When the scenes are ready, click **Open in editor** to edit their text, pictures, narration and timing.
5. To build every scene yourself, click **Create flow** or **Start blank** instead.

### Rehearse before you publish

1. In the editor, click **Rehearse** to hear your avatar narrate the flow in its own voice. Nothing is published.
2. Play one scene, or **Rehearse all** to play every scene in order. A scene's voice is recorded the first time you play it.
3. In rehearsal the avatar speaks over its portrait; the lip-synced video is made when you publish. Lines you have already rehearsed aren't recorded or charged again when you publish.

### Publish a Live Room

1. In the editor, click **Create Live Room**.
2. Choose the room's **Language**, and under **Moving between scenes** choose **Move on automatically** or **Wait for the visitor**.
3. Optional: under **Who this link is for**, add a **Recipient's name** so the avatar greets them by name, and **Notes for your avatar** on how to talk with them. An **Access code** makes visitors enter a code before the room opens.
4. Click **Generate live URL**, then **Copy link** and share it. The link works at once; voices fade in as their narration finishes.
5. Your rooms are under **Live Rooms** on the dashboard, each with its report: visits, conversations, time spent and what people asked.

### Teach your avatar what to answer

1. Open the avatar from **Avatars** on the dashboard, or select it in the editor and click **Edit avatar**.
2. **Purpose**: say what the avatar is there to do in every room, like answering questions about your apartments or teaching your course.
3. **Knowledge**: write its **Background** and short **Facts to know**, and add files or web pages for it to answer from.
4. **Training**: write questions and answers with **Add question**. Under **Visitor questions** you see what visitors asked it in your Live Rooms and what it said; click **Write the answer**, then **Save as Q&A**.
5. The avatar uses all of this when visitors ask questions, in every flow and Live Room it presents, from the next conversation on, without publishing again. Its narration still comes from each flow's script.

### Edit an avatar's voices

1. Open the avatar from **Avatars** on the dashboard, or select it in the editor and click **Edit avatar**. Go to **Voices**.
2. **Change** on the **Default** voice replaces it for every language that has no voice of its own.
3. **Add a language** gives one language its own voice, recorded or picked, in narration and in live answers. **Remove** sends that language back to the default voice.
4. A recorded voice shows **cloning** until it is ready. Until then the default voice speaks that language.

### Change how an avatar appears

1. **How this avatar appears** offers Video avatar (the default: your photo as a talking head), Picture, Animated face, Glow and Voice only. A mascot appears as its own character or as Voice only.
2. On the avatar's page, under **Appearance**, it sets the look that new flows start with. Flows that already have this avatar keep their own look.
3. In the editor, select the avatar on the stage. **Look · this flow** changes it on every scene of this flow; **Look · this scene only** changes one scene.
4. Live Rooms show the new look after you publish the flow again.

### Change the avatar in a flow

1. In the editor, select the avatar on the stage and click **Change avatar**.
2. Choose the avatar to change to. It takes the old avatar's place, and the old avatar's lines move to it, spoken in its own voice.
3. Click **Change on this scene**, or change it on every scene of the flow at once.
4. Publish again: Live Rooms keep the old avatar until you do.

### Use a new voice in a Live Room

1. A Live Room keeps the voices it was published with, so a voice you change on the avatar's page doesn't reach rooms that are already live.
2. Open the room's flow in the editor and click **Publish**. The narration is recorded again in the new voice, so publishing can cost credits; you confirm before anything is charged.
3. Once publishing finishes, every Live Room of that flow speaks in the new voice.

## Virtual, with human standards

Your face and voice are used only with your consent, never anyone else's. Every avatar is disclosed as AI to everyone who visits. It answers from what you gave it, under the rules you set.
