Enterprise Documentation

AVA Console

AVA Console is a full-featured enterprise platform for creating, deploying, and managing AI assistants (personas). From voice interactions to live chat, outbound call campaigns to embeddable widgets — everything you need is in one place.

AI Assistants

Voice Cloning

Call Campaigns

Knowledge Bases

What can you do with AVA Console?

Build AI Assistants

Create custom AI personas with unique names, personalities, system prompts, and knowledge. Each assistant can be embedded anywhere.

Voice Interactions

Enable voice-powered conversations. Clone voices, configure TTS, and let users talk to your AI in real time.

Live Chat Interface

Full-featured chat UI supporting text, voice input, image uploads, file attachments, and streaming AI responses.

Knowledge Bases

Upload documents to knowledge stores and link them to assistants so they answer questions from your own data.

Outbound Call Campaigns

Launch AI-powered outbound phone campaigns, upload contact lists, schedule calls and track results.

Deep Analytics

View full chat history, sentiment analysis, response times, costs, and user engagement metrics.

Embeddable Widgets

Embed a chat or voice widget on any website with a single script tag. Fully customizable appearance.

Plugin Ecosystem

Extend assistants with plugins — connect external APIs, tools, and services to give AI superpowers.

Navigation Overview

Section	Route	Purpose
Dashboard	/dashboard	Overview stats, active assistants, usage summary
Assistants	/dashboard/assistants	Create, edit, delete, embed AI personas
Voices	/dashboard/voice	Manage voice clones and TTS voices
Chat	/dashboard/chat	Interact with assistants via text/voice/file
Knowledge Base	/dashboard/knowledgeStore	Create KB stores, upload and manage documents
Plugins	/dashboard/plugin	Browse, create, configure plugins
Call Campaigns	/dashboard/campaigns/*	Setup and manage outbound AI call campaigns
Analytics	/dashboard/analytics	Chat sessions, costs, sentiment, and insights
SDK & Developer	/dashboard/sdk	Python/JS SDK, API reference, live code runner
API Keys	/dashboard/api_key	Generate, manage, and monitor API keys
Billing	/dashboard/billing	Credits, usage, recharge, transaction history
Custom Domains	/dashboard/customDomains	Add branded domains with CNAME verification
Sub-Users	/dashboard/members	Invite team members, set permissions and credits
Brand Console	/dashboard/whitelabel	White-label your AVA Console with custom branding
Profile	/dashboard/profile	Update personal info, manage workspace members

Dashboard

Your real-time command center. The Dashboard gives you an at-a-glance view of your entire AI operation.

Key Metrics Cards

Current Balance

Your available credit balance shown front and center so you always know your spending runway.

Total Users

Number of end-users who have interacted with your assistants.

Activity Graph

Interactive line/bar chart showing usage trends over time — conversations, tokens, and costs.

Active Assistants

A live grid of your top deployed assistants with their avatars and quick-access status.

Dashboard Sections

Overview Statistics

See your total balance, active users, and monthly spending in the summary cards at the top of the page.

Activity Overview Chart

The interactive chart visualizes conversation volume and token usage. Hover over data points to see daily breakdowns.

Active Assistants Panel

Your most recently used assistants are displayed as cards with avatars. Click any assistant to jump directly to its chat interface.

Quick Actions

From the dashboard you can navigate to any feature in one click. The sidebar remains accessible at all times.

The Dashboard auto-refreshes on account switch. If you manage multiple accounts, switching via the Account Switcher will reload all metrics for the selected workspace.

Assistants

Assistants (also called Personas) are the heart of AVA Console. Each assistant is an AI agent you configure to serve a specific purpose.

Creating an Assistant

Click '+ New Assistant'

Press the Create button in the top-right of the Assistants page to open the creation modal.

Set Name & Title

Give your assistant a unique name (e.g., 'Support Bot') and a descriptive title that appears to end-users.

Upload an Avatar

Upload a profile image for your assistant. This image will appear in chat bubbles and on the embeddable widget.

Write the System Prompt

Define the assistant's personality, role, knowledge scope, and response style. This is the most important configuration step.

Use specific, detailed prompts. Mention the assistant's name, tone (formal/casual), and what topics it should or should not discuss.

Link a Knowledge Base (Optional)

Attach a Knowledge Base to give the assistant access to your custom documents and data.

Enable Voice (Optional)

Toggle voice to let users speak to this assistant. Select a voice from your Voice library.

Save & Deploy

Click Create — your assistant is instantly live and ready for testing via the Chat page.

Managing Assistants

Search & Filter

Quickly find assistants by name using the search bar. Filter by status or voice-enabled state.

Edit Assistant

Update name, avatar, system prompt, linked KB, voice, or any other setting at any time without downtime.

Delete Assistant

Permanently remove an assistant after confirming the deletion prompt. This action is irreversible.

Share Chat Link

Generate a public shareable link for users to chat with this assistant without logging in.

Embed Widget

Get embeddable iframe code or script tag to add this assistant as a chat widget on any website.

Preview Widget

Open a live preview of what the embedded chat widget will look like before publishing it.

Assistant Card Actions

Each assistant card in the grid provides a hover action menu with: Chat (open conversation), Edit, Share (public link), Embed (widget code), and Delete. Owner and Editor roles can edit; Viewer roles can only chat.

Pagination & Search

The assistant grid loads 24 cards per page. Use the search bar for real-time filtered results across all your assistants. The search is debounced and calls the API only after you stop typing.

Voices

Manage the voice library for your AI assistants. AVA Console supports both pre-built voices and custom voice clones.

Browse Voice Library

View all available TTS voices with name, gender, language, and accent. Preview each voice before selecting.

Clone Your VoicePremium

Upload a voice sample to create a custom voice clone. Your clone will appear in the library for use in any assistant.

Preview Voices

Click the play button on any voice card to hear a sample before assigning it to an assistant.

Delete Voice Clones

Remove custom voice clones you no longer need. Only owner-created clones can be deleted.

Adding a Standard Voice

Open Voices Page

Navigate to Voices from the sidebar.

Click '+ Add Voice'

Opens the voice browser modal showing all available system voices.

Filter & Preview

Use the search and language filters. Click the play icon to hear a sample of any voice.

Add to Library

Click Add to add the voice to your library. It will now appear as an option when configuring assistants.

Cloning a Custom Voice

Click '+ Clone Voice'

Opens the custom voice cloning modal.

Enter Voice Name

Provide a descriptive name for the voice clone (e.g., 'John - Sales Voice').

Upload Audio Sample

Upload a clear audio file (MP3/WAV) of the voice you want to clone. Longer and cleaner samples yield better results.

Use a minimum of 30 seconds of clean speech with no background noise for best voice quality.

Submit & Wait

Processing may take a few moments. The cloned voice will appear in your library once ready.

Voice cloning is a premium feature that consumes credits. Ensure you have permission to clone the voice you upload. Only use your own voice or voices you have explicit authorization to clone.

Chat

The Chat interface lets you interact with your AI assistants in real time. It supports text, voice, images, and file uploads — all in one conversation view.

Text Chat

Send text messages and receive streaming AI responses with real-time token-by-token display.

Voice Input

Click the microphone to dictate your message via speech-to-text. The transcript appears in the input field.

Image Upload

Attach images to your message for vision-capable assistants to analyze and respond to.

File Attachment

Upload documents (PDF, DOCX, TXT) that the assistant can reference in its responses.

Knowledge Base Selector

Switch which Knowledge Base the assistant uses mid-conversation without leaving the chat.

Voice Output

Enable TTS to have the assistant's responses read aloud via WebRTC-powered real-time audio.

Streaming Responses

Responses stream token by token for a fluid, real-time feel. No waiting for the full response.

Like / Dislike Feedback

Rate individual AI responses with thumbs up or down to collect quality feedback for your assistants.

Starting a Chat Session

Select an Assistant

Use the assistant selector dropdown at the top of the chat page to pick which persona you want to talk to.

Optionally Link a Knowledge Base

Click the knowledge base selector to attach a KB store for this conversation thread.

Type or Speak Your Message

Use the text input field or click the microphone icon to dictate using your device's microphone.

Send & Read the Response

Press Enter or the send button. The response streams back in real time. Code blocks are syntax-highlighted.

Copy, Like, or Expand

Hover over any message to copy it, rate it, or expand it to full screen for better readability.

Chat Features

You can open the chat in a floating panel mode by clicking the expand icon — useful for multitasking while monitoring AI responses. The chat history persists within the session and can be reviewed in the Analytics section.

If the assistant has voice enabled, toggle the speaker icon to activate real-time voice output. Voice playback uses WebRTC for low-latency audio streaming.

Knowledge Base

Knowledge Bases (KBs) let you store your documents, FAQs, and data so your AI assistants can answer questions based on your actual content.

Create Knowledge Base

Create named knowledge stores with descriptions. Each KB is an isolated container for your documents.

Upload Documents

Upload PDFs, Word docs, plain text, CSV files and more. The platform indexes the content for AI retrieval.

Search Within KB

Search your uploaded documents directly from the KB panel to verify content before linking to an assistant.

Delete Files & KBs

Remove individual files from a KB or delete an entire knowledge base when it's no longer needed.

Link to Assistants

Each KB can be linked to multiple assistants. The assistant view shows which personas use each KB.

Usage Tracking

See how many assistants are using each KB and the total storage consumed.

Creating & Populating a Knowledge Base

Open Knowledge Base page

Navigate to Knowledge Base from the sidebar.

Click '+ New Knowledge Base'

Enter a name and optional description for your knowledge store, then confirm.

Open the KB panel

Click on your newly created KB card to open the document management panel on the right.

Upload your documents

Drag and drop or click to upload files. Supported formats: PDF, DOCX, TXT, CSV, XLSX, and more.

Organize different topics into separate KBs for cleaner retrieval. For example, keep 'Product Manual' and 'HR Policy' in different KBs.

Link to an Assistant

Go to your Assistant settings and select this KB from the Knowledge Base dropdown. The assistant will now retrieve answers from your documents.

Test retrieval in Chat

Start a chat with the assistant and ask questions related to your uploaded documents. The assistant should cite information from your files.

Files are processed asynchronously. Large documents may take a minute to fully index. The file list shows a processing status indicator — wait for it to show 'Ready' before testing.

Plugins

Plugins extend your AI assistants with external tools and capabilities — from web search and weather APIs to custom business integrations.

Create Plugin

Define a custom plugin with name, description, endpoint URL, HTTP method, authentication, and parameter schema.

Edit Plugin

Modify plugin configuration, parameters, authentication headers, or the logo at any time.

View Plugin Details

Inspect the full configuration of any plugin including its API spec, parameters, and last-used timestamp.

Delete Plugin

Remove plugins that are no longer in use. This also detaches them from any assistants using them.

Search & Filter

Search plugins by name or filter by category: Entertainment, Media, Productivity, Device, Settings, and more.

Plugin Logos

Upload custom logos for your plugins to make them visually identifiable in the plugin browser.

Creating a Plugin

Click '+ Add Plugin'

Opens the plugin creation modal.

Fill in the Basic Info

Provide the plugin name, description, and select a category (Entertainment, Productivity, Media, etc.).

Configure the Endpoint

Enter the API endpoint URL and select the HTTP method (GET, POST, PUT, DELETE).

Define Parameters

List the parameters the API expects — name, type, whether required, and a description.

Clear parameter descriptions help the AI know when and how to call the plugin correctly.

Set Authentication

If the API requires authentication, configure the auth type (Bearer, API Key, Basic Auth) and provide credentials.

Upload a Logo

Upload a square image as the plugin's logo. This appears in the plugin browser and assistant settings.

Save & Attach

Save the plugin. Then go to your Assistant settings and attach the plugin to enable it for that AI.

Plugin endpoints are called from the backend. Ensure your API endpoint is publicly accessible or properly authenticated. Sensitive credentials are encrypted at rest.

Call Campaigns

Launch AI-powered outbound voice call campaigns. Your AI assistants make automated calls from a list of contacts — ideal for sales, reminders, surveys, and notifications.

Campaign Sections

Setup Campaign

Create new campaigns, define the assistant, upload contact lists, and schedule call timing.

Manage Campaigns

View running and completed campaigns, see call outcomes, and pause or stop active campaigns.

Manage Numbers

Manage outbound phone numbers used for campaigns. Provision new numbers or configure existing ones.

Setting Up a Campaign

Navigate to Setup Campaign

Go to Run Call Campaign → Setup Campaign from the sidebar.

Create a New Campaign

Click the create button and provide a campaign name and description.

Select an AI Assistant

Choose which AI persona will be making and conducting the calls.

Upload Contact List

Upload a CSV or Excel file with contact names and phone numbers. The platform auto-maps columns.

Ensure phone numbers include country codes (e.g., +1 for US). Invalid numbers are skipped automatically.

Configure Outbound Number

Select an outbound phone number from your provisioned numbers in Manage Numbers.

Set Schedule (Optional)

Choose call timing windows — days of the week and time range — to respect contact availability.

Launch Campaign

Review the summary and click Launch. The campaign starts making calls according to your schedule.

Managing Campaigns

Monitor Live Campaigns

The manage page shows active campaigns with live call counts, success rates, and current status.

View Call History

Drill into any campaign to see the full log of calls — who was called, outcome, duration, and transcript.

Pause or Stop

Pause a campaign to temporarily halt calls, or stop it permanently. Stopped campaigns cannot be restarted.

Outbound call campaigns are subject to carrier regulations and compliance rules. Ensure you have consent from contacts before calling and comply with local telemarketing laws (TCPA, GDPR, etc.).

Analytics

Gain deep visibility into every conversation your AI assistants have. Track performance, costs, user sentiment, and engagement across all channels.

Chat Session Viewer

Browse all chat sessions with full conversation transcripts. Filter by assistant, date range, or user.

Usage Metrics

Track total conversations, token usage, cost per session, and API call volume over time.

Response Analytics

See average response times, streaming latency, and quality scores per assistant.

Export to Excel

Download filtered session data as an XLSX file for offline reporting or further analysis.

Advanced Filtering

Filter sessions by assistant, date range, session type (chat/voice), cost range, and more.

Full Conversation View

Click any session to read the complete message thread with sender labels, timestamps, and code rendering.

Using Analytics

Open Analytics Page

Navigate to Analytics from the sidebar.

Select an Assistant (Optional)

Use the assistant filter to scope analytics to one specific persona, or view all assistants together.

Apply Date Filters

Set a start and end date to zoom into a specific time period. The session list updates in real time.

Browse Sessions

The session list shows each conversation with the assistant name, start time, message count, and estimated cost.

Open a Session

Click any session to expand the full conversation panel. Messages are rendered with markdown and code highlighting.

Export Data

Click the Export button to download the current filtered view as an Excel file for reporting.

Session data is retained indefinitely. You can always go back and review historical conversations to audit AI responses or identify common user questions for improving your system prompts.

SDK & Developer

Integrate AVA Console's AI capabilities directly into your own applications using our Python and JavaScript SDKs, or by calling the REST API directly.

Python SDK

Full-featured Python library for chat completions, streaming, knowledge base queries, and voice synthesis.

JavaScript SDK

Node.js and browser-compatible JS SDK with identical feature parity to the Python SDK.

Live Code RunnerInteractive

Execute SDK code examples directly in the browser without any local setup. See real API responses instantly.

API Reference

Complete endpoint documentation with request/response schemas, authentication details, and error codes.

Code Examples by Category

Organized code examples covering: Chat, Streaming, Knowledge Base, Voice, Campaigns, and Webhooks.

API Key Integration

Your API key is auto-populated in code examples once you generate one from the API Keys section.

Quick Start — Python

bash

pip install ava-ai-sdk

python

from ava_sdk import AvaClient

client = AvaClient(api_key="YOUR_API_KEY")

# Send a message to an assistant
response = client.chat(
    persona_id="YOUR_PERSONA_ID",
    message="Hello! How can you help me today?"
)
print(response.text)

Quick Start — JavaScript

bash

npm install ava-ai-sdk

javascript

import { AvaClient } from 'ava-ai-sdk';

const client = new AvaClient({ apiKey: 'YOUR_API_KEY' });

// Stream a response
const stream = await client.chat.stream({
  personaId: 'YOUR_PERSONA_ID',
  message: 'Tell me about your capabilities.',
});

for await (const chunk of stream) {
  process.stdout.write(chunk.text);
}

SDK Page Features

Select Language

Toggle between Python and JavaScript using the language tabs at the top of the SDK page.

Browse Example Categories

Examples are organized by category: Overview, Chat, Streaming, Knowledge Base, Voice, and more.

Edit & Run Code

Modify any code example directly in the browser editor, then click Run to execute it against live APIs.

View API Reference

Switch to the API Reference tab for full REST endpoint documentation with curl examples.

API Keys

Generate and manage API keys for programmatic access to AVA Console's AI capabilities from your own applications and services.

Generate API KeyCopy Once

Create new API keys with custom names. Each key is shown once at creation — copy it immediately.

Usage Metrics

See per-key usage stats: total requests, token consumption, cost, and last-used timestamp.

Usage Charts

Interactive bar and line charts showing daily/weekly API call volume and cost breakdown per key.

Search Keys

Search across your API keys by name to quickly locate a specific key in large organizations.

Revoke KeysIrreversible

Delete any API key to immediately revoke access. Revoked keys become permanently invalid.

Secure Storage

API keys are hashed server-side. Even AVA cannot retrieve a key after it's been created — only you see it once.

Creating an API Key

Open API Keys page

Navigate to API Keys from the sidebar.

Click '+ Generate Key'

Opens the key creation dialog.

Name your key

Give the key a descriptive name that identifies its intended use (e.g., 'Production Backend', 'Mobile App').

Copy the key immediately

The full API key is shown exactly once. Copy it to a secure location — you cannot view it again.

Store your API key in an environment variable or secrets manager, never hardcode it in source code.

Use in your application

Pass the key as a Bearer token in the Authorization header of all API requests.

Never expose your API key in client-side code, public repositories, or unencrypted communications. Treat it like a password.

Billing

Manage your credits, track usage costs, and recharge your account. AVA Console uses a credit-based consumption model.

Current Balance

See your live credit balance, how much you've consumed this billing period, and a progress bar toward depletion.

Recharge Credits

Top up your credit balance instantly using the Recharge modal. Multiple denomination options available.

Usage Charts

Line and bar charts show credit consumption over time — broken down by chat, voice, and API usage.

Transaction History

Full list of all credit purchases, API usage deductions, and adjustments with timestamps.

Usage Breakdown

Detailed breakdown of how credits are consumed: chat tokens, TTS characters, voice minutes, and API calls.

Sub-User Credit Allocation

If you're a sub-user, your billing page shows only your allocated credits and usage limits set by the owner.

Recharging Credits

Click 'Add Credits'

Opens the Recharge modal with available credit packages.

Select a Package

Choose a credit denomination that suits your usage needs. Larger packages typically offer better value.

Complete Payment

Follow the payment flow. Credits are added to your balance immediately upon successful payment.

Verify Balance

Return to the Billing page to confirm the updated balance in the balance card.

Credit consumption rates vary by feature. Chat messages cost per token, TTS costs per character, voice cloning costs per processing job, and outbound calls cost per minute. Check the pricing page for current rates.

Custom Domains

Map your own branded domain to AVA Console so your users interact with your AI at your URL — not ava.pathor.ai.

Add Custom Domain

CNAME Verification

Verify ownership by adding a CNAME record in your DNS provider pointing to console.ava.pathor.ai.

Domain Status Tracking

Monitor domain verification status: Pending, Active, or Suspended — with real-time updates.

Remove Domain

Delete a domain mapping to free it from your account and remove the DNS routing.

Setting Up a Custom Domain

Go to Custom Domains

Navigate to Settings → Custom Domains from the sidebar.

Click '+ Add Domain'

Enter your custom domain or subdomain (e.g., assistant.mycompany.com) and submit.

Add CNAME Record

Log into your domain registrar/DNS provider. Add a CNAME record pointing your domain to:

Use a subdomain (e.g., ai.company.com) rather than a root domain — CNAME records are not supported on root domains by most DNS providers.

dns

Type:   CNAME
Host:   ai (or your chosen subdomain)
Value:  console.ava.pathor.ai
TTL:    3600

Wait for Propagation

DNS changes can take up to 48 hours to propagate globally, though most changes resolve within 1 hour.

Verify in Dashboard

Return to the Custom Domains page and click Verify. Once DNS resolves, the status changes to Active.

Configure White-Label (Optional)

Go to Brand Console to set colors, logos, and a console name for your custom domain.

Sub-Users & Members

Collaborate with your team by inviting sub-users. Control exactly what each member can access, how many credits they can use, and what actions they're allowed to perform.

Invite Sub-Users

Send email invitations to team members. They receive a setup link to join your AVA Console workspace.

Role-Based Permissions

Assign Admin, Editor, or Viewer roles. Each role has different levels of access to features.

Granular Permissions

Fine-tune access per feature module: Assistants, Chat, Voices, Knowledge Base, Analytics, Billing, etc.

Credit Allocation

Allocate a credit budget to each sub-user. They can only spend up to their assigned limit.

Usage Limits

Set monthly message limits and API call quotas for each team member to control costs.

Block / Unblock

Instantly block a member to suspend all their access without removing them from the workspace.

Inviting a Team Member

Navigate to Sub-Users

Go to Settings → Sub-Users from the sidebar.

Click '+ Invite Member'

Opens the invitation dialog.

Enter email address

Type the member's email. They will receive an invitation email with a setup link.

Assign a Role

Select Admin (full access), Editor (can create/edit), or Viewer (read-only).

Set Permissions

Toggle individual feature permissions on or off for this member.

Allocate Credits

Set a credit budget that will be drawn from your main balance for this member's usage.

Start with a conservative credit limit and increase it as you gauge the member's usage patterns.

Send Invite

Click Send. The member receives an email and can log in once they accept the invitation.

Roles Reference

Role	Create/Edit	Delete	View Analytics	Billing Access	Invite Members
Owner	✅	✅	✅	✅	✅
Admin	✅	✅	✅	Limited	✅
Editor	✅	❌	✅	❌	❌
Viewer	❌	❌	✅	❌	❌

Brand Console

White-label AVA Console with your own brand identity. Replace AVA's logo, colors, and name with your company's branding.

Logo Upload

Replace AVA's default logo with your company's logo. Shown in the sidebar, login page, and embeds.

Brand Colors

Customize primary, secondary, and accent colors to match your brand's color palette.

Console Name

Replace 'AVA Console' with your product name throughout the interface.

Enable Branded Console

Toggle to activate your branding across all pages when users log in on your custom domain.

Live Preview

See a real-time preview of how your branding will look before going live.

Domain Scoped

Branding only activates for users visiting via your custom domain, keeping the main AVA Console unchanged.

Configuring Your Brand

Open Brand Console

Navigate to Settings → Brand Console from the sidebar.

Upload your logo

Click the logo upload area and select your company logo (PNG or SVG recommended, square format).

Set brand colors

Pick your primary and accent colors using the color pickers. Enter hex values for precision.

Enter your console name

Type your product or company name to replace 'AVA Console' in the interface.

Enable the branded console

Toggle 'Enable Brand Console' to activate your branding. Users on your custom domain see your brand.

Ensure you have a Custom Domain set up and verified before enabling the brand console for the best experience.

Save changes

Click Save. Changes take effect immediately for all users on your custom domain.

Brand Console is available to workspace owners only. Sub-users and team members cannot modify branding settings.

Profile

Manage your personal account information, profile picture, and workspace team members.

Update Profile

Edit your first name, last name, and email. Changes are saved to your account immediately.

Profile Picture

Upload a profile picture that appears in the sidebar user card and throughout the interface.

Team Member Management

Invite, manage roles, and remove workspace team members (separate from sub-users).

Invite to Workspace

Send email invitations to collaborators to join your workspace with a specific role.

Updating Your Profile

Go to Profile

Click your avatar in the bottom-left corner of the sidebar, then select Profile.

Update your details

Edit your first name, last name, or email address in the profile form.

Upload a photo

Click the camera icon on your avatar to upload a new profile picture.

Save changes

Click Save to persist your profile updates.

Profile page has two tabs: Profile for personal info and Team for managing workspace collaborators with role assignments (Admin, Editor, Viewer).

Embeddable Widgets

Add an AI chat or voice widget to any website in minutes. Just paste a script tag — no framework required.

Chat Widget

Embeddable floating chat button that opens a full AI chat panel. Customize colors, position, and features.

Voice Widget

Voice-only embeddable widget for hands-free AI interaction. Click to speak, AI responds in audio.

Full Customization

Control icon color, header background, button position, enable/disable text vs. speech, and more via data attributes.

Works Everywhere

IIFE-based scripts work on any website regardless of framework — React, Vue, plain HTML, WordPress, etc.

Widget Preview

Use the Widget Visual Playground at /dashboard/widget-preview to see live changes before deploying.

API Key Authentication

Widgets authenticate via your API key and persona ID passed as data attributes in the script tag.

Embedding the Chat Widget

Open Assistant Embed modal

Go to Assistants, hover over your assistant card, and click the Embed icon.

Copy the embed code

The modal shows a pre-configured script tag with your API key and persona ID already filled in.

Paste into your website

Add the script tag to your HTML, just before the closing </body> tag.

html

<script
  src="https://console.ava.pathor.ai/scripts/chat-widget.js"
  data-api-key="YOUR_API_KEY"
  data-persona-id="YOUR_PERSONA_ID"
  data-enable-text="true"
  data-enable-speech="true"
  data-position="bottom-right"
  data-icon-color="#6366f1"
  data-header-bg-color="#4f46e5"
></script>

Widget Configuration Options

Attribute	Required	Default	Description
data-api-key	Yes	—	Your AVA Console API key
data-persona-id	Yes	—	The ID of the assistant persona
data-enable-text	No	true	Enable text chat input
data-enable-speech	No	false	Enable voice input/output
data-position	No	bottom-right	Widget position: bottom-right, bottom-left
data-icon-color	No	#6366f1	Widget launcher button background color
data-header-bg-color	No	#4f46e5	Chat panel header background color

Embedding the Voice Widget

html

<script
  src="https://console.ava.pathor.ai/scripts/voice-widget.js"
  data-api-key="YOUR_API_KEY"
  data-persona-id="YOUR_PERSONA_ID"
  data-position="bottom-right"
  data-icon-color="#6366f1"
></script>

Use the Widget Visual Playground at /dashboard/widget-preview to tweak all appearance settings interactively and see live changes before committing to an embed code.