10,000 listeners in 30 days. AI hosts, real guests.Read the field note
DocsUpdated May 13, 2026

Talent profile reference

Field-by-field reference for every section of a talent profile page in Mato.

10 min read
1815 words
talent
profile
personality

This reference covers every section on a talent profile page. Use it to understand what each field does, where data comes from, and how sections relate to each other.

To open a talent profile, go to Talent in the sidebar and click any host name.

Hero card

The hero card sits at the top of the profile and combines identity information with the personality radar chart.

Portrait

The left side of the hero card shows the talent's portrait image. Mato generates this automatically when you create a host. While the image is being generated, you see a pulsing placeholder with a spinner. If generation failed, an error icon appears instead.

Click the portrait to open the avatar overlay, which lets you regenerate the image or view the prompt used to create it.

Identity fields

These fields appear in the hero card next to the portrait.

FieldDescription
NameDisplay name for the host. Shown in scripts, audio credits, and the talent directory.
Host type badgeEither AI Character or Real Person. Real Person is currently marked as Coming Soon.
Gender badgeMale, Female, or Non-binary. Only shown if set to something other than "Unspecified".
TaglineA one-line summary of the host's angle or persona. Falls back to the short bio or personality summary if no tagline is set.
Short bioA sentence or two describing the host. Appears below the tagline when both are present.
Created dateRelative timestamp ("3 days ago", "2 months ago") showing when the talent was added.
VoiceName of the assigned voice (Gemini or ElevenLabs). Hidden if no voice is configured.
Hosting styleA short phrase describing the host's delivery approach (e.g., "Conversational storyteller" or "Data-driven analyst"). Appears in the talent picker when assigning hosts to a podcast.
Linked podcastsCount of podcasts this talent is assigned to.

Radar chart

The radar chart appears on the right side of the hero card. It plots five personality dimensions on a pentagon:

AxisWhat it measures
ImproviseHow much the host goes off-script, riffs, and ad-libs vs. sticking to the written script.
ExplorerCuriosity level. High scores mean the host digs into tangents and follows up on side threads.
HumorFrequency and weight of jokes, wordplay, and lighthearted moments in delivery.
FormalRegister and tone. High formal means polished, measured language. Low formal means casual, colloquial speech.
TechnicalDepth of domain-specific language. High technical means the host uses jargon and assumes audience expertise.

Each axis runs from 0% to 100%. The shaded polygon shows the host's position on all five at once.

Tags

Below the radar chart, a row of tags summarizes the host's personality in short keywords (for example, "curious", "witty", "data-driven"). Tags are generated along with the spec grid but can be edited manually on the edit page.

Personality narrative

This card contains a free-text description of the host's personality written in paragraph form. It gives the AI script generator context about how the host thinks, reacts, and structures arguments. The narrative is longer and more detailed than the tagline or short bio.

If no narrative has been written, this card is hidden.

Characteristics

Characteristics are label/value pairs that capture specific personality traits, communication habits, and behavioral patterns. Each card in the grid shows:

  • Label (uppercase, muted) such as "Pace", "Vocabulary", or "Conflict Style"
  • Value (body text) describing the trait, for example "Tends to speed up when excited, slows down for emphasis"

The grid displays up to six characteristics by default. If the talent has more than six, a Show all button expands the list. Characteristics are editable on the talent edit page, with a maximum of 12 per profile.

Sample conversations

This section shows a simulated back-and-forth between the host and a co-host. Each exchange displays:

  • Speaker name in small muted text above the bubble
  • Dialogue text in a chat-style bubble

The host's lines appear on the left with a muted background. The co-host's lines appear on the right with a bordered background. This section helps you preview the host's tone, vocabulary, and conversational rhythm before assigning them to a podcast.

Sample conversations are generated as part of the spec grid and can hold up to eight exchanges. If no conversation data exists, this section is hidden.

Best-for cards

Best-for cards describe the types of shows, topics, and formats where this host works well. Each card contains:

ElementDescription
IconAutomatically selected based on keywords in the title (news, interview, tech, ethics, and so on). Defaults to a sparkle icon if no keyword matches.
TitleShort name for the use case, like "Breaking News Roundups" or "Deep-Dive Interviews".
DescriptionA sentence explaining why the host fits this format.

Cards are laid out in a two-column grid. If no best-for data exists, this section is hidden.

FAQ

The FAQ section uses an accordion layout. Each entry has a question (shown as the collapsed header) and an answer (shown when expanded). The first two entries are expanded by default.

FAQ entries are generated alongside the spec grid. They typically cover questions like "How does this host handle disagreements?" or "What topics does this host avoid?" You can edit, add, or remove entries on the edit page.

If no FAQ entries exist, this section is hidden.

Audio samples

Audio samples let you hear how the talent sounds with their current voice configuration. Each sample shows:

  • A label badge identifying the clip (for example, "Intro", "Debate", "Story")
  • A duration in seconds
  • A built-in audio player with playback controls

Samples are generated from the talent edit page (not the profile page). While generation is in progress, the card shows a spinner. If generation fails, an error message appears with guidance to check the voice configuration.

The audio samples card is hidden when there are no samples and no generation is in progress.

Personality data

The personality data card shows the structured personality traits stored for this host. These come from either the creation wizard questionnaire (for hosts built from scratch) or from the template that was forked.

The card organizes data into sections:

  • Traits (tone, vocabulary, perspective, and so on)
  • Interview insights if the host was created via questionnaire (identity, communication style, speaking patterns, opinions, experience, podcast context)
  • Raw answers completion status showing how many questionnaire fields were filled in

If no personality data exists, the card shows an empty state with a link to add traits from the edit page.

An Edit button in the card header takes you to the talent edit page.

Voice identity

The voice identity card controls how the talent sounds during audio rendering. It holds these settings:

SettingDescription
Gemini voiceThe primary TTS voice. Pick from the Gemini voice library, filtered by gender and style tags.
ElevenLabs voiceSecondary TTS voice, available to select accounts. Overrides Gemini when set.
AccentEnglish accent variant (American, British, Australian, and others).
Speech speedPacing preset: Slow, Normal, Fast, or Very Fast.
Audio profileFree-text description of how the talent should sound in the mix.
Director notesFree-text delivery guidance for tone, pacing, and emphasis.

Changes to voice settings are saved independently from other profile data. The save button stays disabled until at least one voice (Gemini or ElevenLabs) is selected.

For a full walkthrough of each voice setting, see Configure voice and delivery.

Memory

The memory card shows what the AI agent remembers about this host across episodes. It has four tabs:

TabContents
IdentityThe host's core persona text, describing who they are and how they present themselves.
StyleCommunication and delivery style preferences the agent has learned.
EpisodesA log of episodes the host has appeared in, with titles, dates, and topic tags.
StoriesAnecdotes, quotes, and personal stories the host has used, with source context.

Memory is powered by Mato's AI memory system. If the talent does not have an active memory agent, the card shows an empty state explaining that memory is not yet available.

The card automatically loads memory data when an agent is present. If the load fails, a retry button appears.

Agent status

The agent status card tracks the sync state between the talent profile and its memory agent. It displays:

ElementDescription
Status badgeColor-coded: green for Synced, yellow for Pending, red for Failed or Error.
Last syncedRelative timestamp of the most recent sync. Shows "Never" if the agent has not synced yet.
Sync Now buttonManually triggers a sync between the talent profile and the memory agent.

Linked podcasts

The linked podcasts card lists every podcast this talent is assigned to. Each row shows:

  • Podcast name
  • Position (the host order in the podcast)
  • Role (the host's function in the show)
  • A link to Open Settings which navigates to the podcast's hosts tab

If the talent is not assigned to any podcast, a dashed empty state appears.

Visibility levels

Visibility controls who can see and use the talent. There are three levels:

LevelWho can see it
PrivateOnly the person who created the talent. Other team members cannot see or use it.
AccountAll members of the workspace (team). This is the default for new talent.
PublicVisible in the talent directory. Anyone browsing the directory can view the profile and fork it into their own workspace. A talent directory guide is planned for a future Help Center update.

You set visibility on the talent edit page. Changing visibility does not affect existing podcast assignments.

Talent types

Mato supports two talent types, selected during creation:

AI Character is a fully fictional host. Mato generates its personality, portrait, and voice from scratch based on a text description or template. All current talent creation flows produce AI Character hosts.

Real Person is designed for hosts based on a real individual, where personality data comes from interviews or questionnaires. This type is marked as Coming Soon in the creation wizard and cannot be selected yet.

The host type badge on the hero card shows which type applies. The type cannot be changed after creation.

Still need help?

If the doc didn't solve it, open Intercom and ask us directly. Include the page you're on and what you expected to happen.

Related guides

Getting started

Create your Mato account, set up your workspace, and get ready to produce your first podcast.

Configure podcast settings

Adjust your podcast's name, schedule, AI behavior, audio, branding, publishing connections, and more from a single settings page.

Create a custom host

Add a new AI host to your workspace by building one from scratch or forking from a template.

Set up an interview

Create a live AI interview, invite a guest, and prepare questions.

Manage your team

Add teammates, assign roles, and control who can do what in your Mato workspace.

Subscribe to RSS feeds

Follow Mato's changelog and blog through RSS. Add the feeds to your preferred reader and get updates as they publish.

Create social clips

Turn your podcast episodes into short-form video clips for TikTok, Instagram Reels, and YouTube Shorts.

Episodes

Generate, edit, and refine podcast episodes inside Mato.

Customize branding and cover art

Set up your podcast's visual identity with AI-generated covers, custom uploads, color schemes, and brand profiles.

Configure voice and delivery

Set a voice, accent, speed, and delivery style on a talent profile so episodes sound exactly how you want.

Guest recording guide

What to expect when you join a Mato interview as a guest.

Invite members to your workspace

Add teammates to your Mato workspace by sending email invitations, choosing roles, and managing pending invites.

Customize clip templates and styles

Choose a visual template for your social clips, adjust caption styling, override colors and fonts, and preview changes before rendering.

Distribution

Connect Mato to your website and podcast apps so episodes can go live with less manual work.

Review interview segments

Score, filter, and select the best interview segments for your episode.

Plans and pricing

A complete reference for Mato's pricing tiers, feature availability, usage limits, and billing management.

Connect social accounts

Link your YouTube, TikTok, Instagram, and LinkedIn accounts so you can publish clips straight from Mato.

Turn an interview into an episode

Generate a full podcast episode from your best interview segments.

Set up a schedule

Configure automated episode generation on a recurring schedule.

Email preferences

Control which emails Mato sends you, turn off individual categories, or unsubscribe from everything.

Publish clips to social media

Post your rendered social clips to LinkedIn, TikTok, Instagram, and YouTube directly from Mato.

Record with video

Enable video capture during interviews to pair visual context with your audio recording.

Use the schedule calendar

View, manage, and rearrange upcoming episodes on the schedule calendar.

Manage episode suggestions

Review, accept, and use AI-generated topic suggestions to create episodes.

Set up Listener Q&A

Enable Listener Q&A so your audience can submit questions through web forms, SMS, voice, or WhatsApp, then configure channels, moderation rules, and notifications.

Manage listener questions

Review incoming listener questions, approve or reject them, feature questions for episodes, and track Q&A performance.

© 2026 Mato. All rights reserved. English · Multiple languages available