Introducing Griffin: 48% thought it was a real person, the first model to pass the real-time video Turing test. Learn more.
Products
build with charlie
PAL Maker
No-code way to build and deploy a PAL
explore cvi
Developer API
Build and deploy PALs with the API 
Enterprise
Enterprise
Book Demo
By Use Case
AI SDR
Healthcare Agents
Interview Agents
L&D Agents
Custom Agents
API PLatform
API PLatform
Get Started
Developer API
Docs
API Reference
Pricing
Features
Overview
Magic Canvas
Presentation
Components
Internet Search
‍Google Meet
Memory
Knowledge Base
Research
Research
Explore
Research Overview
Knowledge Navigator
Models
Phoenix-4.5
Raven-2
Sparrow-2
Griffin (Preview)
Vision
pricing
LoginGet Started
get started
LoginGet Started
Select an Account Type

Choose how you want to experience Tavus. Whether you’re building with our APIs or meeting a PAL, you can switch anytime.

Developer Account

Build real-time, human-like AI experiences using Tavus APIs and tools.

Best for developers, founders, and teams integrating Tavus into a product.

NEW DEVELOPER ACCOUNT
PALs Account

Meet your personal AI companions who listen, remember, and are always present.

Best for individuals looking to talk, explore, and connect with a friend.

NEW PALS ACCOUNT
starter kit

Hi, I’m Jolene from Hypothetical Health

Patient Intake Nurse

Try the demo

Jolene

Medical Intake

Pre-visit intake patients finish in minutes, face to face

A nurse-led intake conversation that collects name, date of birth, reason for visit, history, and pain level by voice, reads it back for confirmation, and sends a concise report to the care team before the patient is called. Ships with a frontend, deploy script, and report webhook.

Use this kit
Try the demo

WHO IT’S FOR

Built for the teams who have this conversation every day.

Clinic operations and digital-health teams at practices, urgent care, and telehealth providers who want the room ready before the patient is in it.

01

A conversation, not a clipboard

Jolene asks the five intake questions out loud and listens. An on-screen box appears only when a spoken answer can't be captured reliably, so patients who struggle with forms don't struggle here.

02

Read back and confirmed by the patient

An intake summary card shows what she heard. The patient confirms or corrects it, and the submit_intake_report tool packages a factual report for the nursing staff. What happens next is the care team's call.

03

Stays in its lane

Jolene never diagnoses, recommends treatment, or draws clinical conclusions, and she pauses the intake if the camera shows a privacy or safety concern. Clinical judgment stays with your team.

WHAT’S INSIDE

Everything a PAL needs for the conversation, already set up.

Medical Intake

Skills

Perception, Magic Canvas

Attached to the PAL and on from the first call.

Tools

submit_intake_report, end_call

What the PAL can do mid-conversation, beyond talking.

Face

Jolene

Keep the face, or pick another from the library.

Models

Phoenix-4.5, Raven-2, Sparrow-2

Face, perception and timing, already wired.

First words

Hi, I’m Jolene from Hypothetical Health

I’m a registered nurse handling patient intake, and I’ll run a short pre-visit check-in and pass a concise report to the nursing staff.

Ships with

Structured intake, Frontend, Deploy script

In the kit the moment you copy it.

Kit type

Full customization

One-click copies straight into PAL Maker. Full customization adds a frontend and deploy script.

Deploy

Hosted page, widget, embed, or Google Meet

Pick where the conversation happens on the Deploy tab.

How a conversation runs

Three beats, start to finish.

THE SAMPLE SCENARIO

How the sample conversation runs

Jolene runs pre-visit patient intake for a clinic. She holds a short, natural conversation and collects five details by voice: full name, date of birth, reason for visit, medical history, and pain level, using an on-screen box only when a spoken answer cannot be captured reliably. She reads the details back on an intake summary card for the patient to confirm, then prepares a concise factual report for the nursing staff. She never diagnoses, recommends treatment, or draws clinical conclusions, and she pauses the intake if the camera shows a privacy or safety concern. Her name, her clinic, and her greeting are sample details; replace them with yours.

BEAT 01

Consent

Set expectations, confirm consent, and collect visit context.

Beat 01 · the PAL opens

Jolene

Jolene

Patient Intake Nurse

Consent

BEAT 02

Intake call

Ask structured symptom, history, medication, and visit-goal questions.

Beat 02 · in the conversation

Jolene

Jolene

Patient Intake Nurse

Intake call

BEAT 03

Care summary

Format intake answers into a reviewable summary for the receiving team.

Beat 03 · the close

Jolene

Jolene

Patient Intake Nurse

Care summary

MAKE IT YOURS

Answer a few questions and PAL Maker does the rest.

Open the kit in PAL Maker and it asks for the details that matter here. Everything else, from the prompt to the face, stays editable.

STEP 01

Answer three questions

Open the kit in PAL Maker and Charlie asks for the details that make it yours. Your answers stand in for the sample scenario.

Charlie asks

What's your clinic, and what kinds of visits does it see?

Which intake details does your care team need?

Where should the intake report be sent?

STEP 02

Test it face to face

Talk to your PAL on the Test tab. Adjust the prompt, swap the face or voice, add knowledge, and try again until it sounds like your team.

Test tab

A live video call with your PAL, before anyone else meets it.

Perception, Magic Canvas and tools all run in the test call.

STEP 03

Deploy where the conversation happens

Go live from the Deploy tab as a hosted page, a widget on your site, an embed in your app, or a teammate in Google Meet.

Deploy tab

Hosted page · Widget · Embed · Google Meet

Share a link, or drop in one script tag.

Use this starter kit
Try the demo

QUESTIONS

Three things people ask before they open a kit.

The short answers. The rest are on the Starter Kits page, under Starter Kit questions, answered.

WHAT IT COSTS

Opening a kit is free

Open any kit in PAL Maker at no cost and talk to it right away. Conversations use the minutes on your plan, and there's a free plan to start on.

YOUR DATA

Conversations stay in your account

Transcripts and recordings belong to your Tavus account, not to the kit. Tavus is SOC 2, GDPR and HIPAA compliant; the trust center has the details.

FACE AND VOICE

Change the face, voice and name

Pick any stock face and voice in PAL Maker, or train your own from a short video or a single image. The conversation the kit ships with stays the same.

MORE STARTER KITS

Not the conversation you had in mind?

STARTER KITS

David

Candidate Interviewer

Interviewer

Deborah

Sales Discovery Call

Sales Coach

Ray

Patient Simulation

Mock Patient

Nina

Tutor

Biology Tutor

Andrea

Employee Onboarding

HR Onboarding Specialist

Nathan

Customer Service Coaching

Customer Service Coach

Browse all kits
Open PAL Maker
Starter Kits

Browse every Starter Kit.

See all kits
Enterprise

Want Tavus to build and run a custom PAL with your team?

TALK TO A (real) HUMAN

company

PricingVisionEnterpriseCareersPartnerships

Resources

BlogPerspectivesBrand kit (download)Press kitInfo for AIs

developers

DocsAPI referencePAL MakerQuickstartllms.txt

research

Phoenix-4.5 (Rendering)Raven-2 (Perception)Sparrow-2 (Turntaking)Griffin (Preview)See all research

socials

LinkedInXDiscord

legal

ADAPrivacy policyTerms of serviceWebsite terms of serviceAcceptable use policy

Support

DiscordEmail support@tavus.ioSupport centerTrust center
explore with ai:
© 2026 Tavus   |  THE HUMAN COMPUTING COMPANY  |  All Rights Reserved