AI Teacher
Explains topics, adapts to learner level, and teaches step by step.
Avatrix is a multi-purpose avatar app that can be integrated into any website with simple changes. It can be used for many service experiences from customer service and reception to tutoring, onboarding, and other interactive workflows.
Avatrix Runtime
Embeddable avatar interface
Host Website
Integrate with simple changes and launch the avatar experience inside your existing UI.
Use Cases
Protected Runtime Access
Runtime starts only after authorized host-to-runtime initialization.
Explains topics, adapts to learner level, and teaches step by step.
Welcomes visitors, collects intent, and routes them to the right next step.
Handles common support questions, troubleshooting, and escalation intake.
Introduces users to a product, service, or organization and helps them get started.
Guides applicants through programs, eligibility, deadlines, and required documents.
Qualifies needs, recommends options, and moves leads toward the right offer.
Avatrix is a business-ready multi-purpose avatar app. It can be embedded into almost any website with minimal changes and adapted to many service flows.
Runtime appears only after a trusted host completes the handshake with AVATRIX_INIT.
Integration note
The avatar runtime can be restricted to approved host websites while the product itself remains easy to embed across your approved properties and workflows.
Service catalog API
Hosts can discover supported roles and accepted parameters from /api/services.
Pricing is made up of three parts: platform access and hosting, AI consumption, and any custom integration or service design. The total depends on deployment scope, traffic, voice activity, model level, and support requirements.
The base cost covers access to the avatar runtime, secure hosting, approved-origin controls, administration, and the configured service framework used by the deployment.
Variable cost is driven by actual AI consumption such as conversation volume, token usage, speech-to-text, text-to-speech, realtime voice minutes, and the model tier selected for the experience.
Project cost also depends on branding, custom roles or workflows, host-app integration, API or LMS connections, analytics, security requirements, and any ongoing support or SLA commitments.
A deployment with higher traffic, more realtime voice, heavier speech generation, or premium model usage will cost more than a low-volume text-focused rollout.
This structure keeps the fixed platform portion clear while aligning the variable portion with the actual resources used by each client deployment.
Integration is simple: embed the runtime, wait for the ready handshake, then send one AVATRIX_INIT message with the host SSO token. That token carries the project API key for organization and project resolution. Runtime options like service, language, and personalization remain optional.
source"avatrix-host"Required message origin marker for the host-to-runtime handshake.
type"AVATRIX_INIT"Required bootstrap event that starts the runtime session.
payload.sso_token"<host token>"Required signed host token. Its payload should carry avatrix_api_secret, optional project_id, plus iat and exp so the iframe can resolve the registered organization and project.
payload.serviceRoleOptional service role selection, such as "AI Teacher" or "AI Receptionist". If omitted, the runtime falls back to AI Teacher.
payload.serviceConfigOptional role settings for brand, goals, welcome style, routing, and scope.
payload.first_nameOptional personalization stored for later use when the host sends CHAT/GREET_USER.
payload.userIdOptional internal identifier kept for runtime context only.
payload.userLanguageOptional locale or language hint for the conversation.
payload.avatarPathOptional avatar override. If omitted, the active service chooses the default avatar.
payload.voice / payload.openaiVoiceOptional speech voice override for the active role.
payload.openaiModelOptional model override when the host needs a different LLM profile.
payload.systemPromptOptional extra runtime instructions layered on top of the selected service.
payload.chatLogOptional prior conversation state to resume an existing session.
payload.backgroundImageOptional runtime background customization for the embed.
This is enough to authenticate the project and start a role-aware runtime.
window.parent.postMessage({
source: "avatrix-host",
type: "AVATRIX_INIT",
payload: {
sso_token: hostSignedToken,
project_id: AVATRIX_PROJECT_ID,
serviceRole: "AI Receptionist",
serviceConfig: {
organizationName: "North Clinic",
welcomeMessage: "Welcome visitors warmly and ask how you can help.",
knowledgeScope: ["appointments", "billing", "visitor check-in"],
escalationTarget: "Front desk staff"
},
first_name: "Amina",
userLanguage: "en"
}
}, AVATRIX_RUNTIME_ORIGIN);
window.parent.postMessage({
source: "avatrix-host",
type: "CHAT/GREET_USER"
}, AVATRIX_RUNTIME_ORIGIN);avatrix_api_secretRequired claim carrying the Avatrix API secret issued for the selected project.
project_idOptional claim used when the host already knows the concrete Avatrix project id.
iat / expRequired issued-at and expiry claims so the signed token remains time-bound.
The host should send the SSO token through postMessage. The iframe decodes the token, extracts the project API key, resolves the registered project, and meters usage against that project.
Allowed origins are configured in the dashboard for each project. The host does not send allowed origins at runtime.
ai_teacherExplains topics, adapts to learner level, and teaches step by step.
receptionistWelcomes visitors, collects intent, and routes them to the right next step.
customer_supportHandles common support questions, troubleshooting, and escalation intake.
onboarding_guideIntroduces users to a product, service, or organization and helps them get started.
admissions_advisorGuides applicants through programs, eligibility, deadlines, and required documents.
sales_advisorQualifies needs, recommends options, and moves leads toward the right offer.
career_coachSupports career planning, interview practice, and skill-growth decisions.
serviceConfig.assistantNamestringOptional persona name spoken or referenced by the avatar.
serviceConfig.organizationNamestringHost brand or organization the avatar represents.
serviceConfig.audiencestringWho the avatar is serving in this embed, such as students or visitors.
serviceConfig.primaryGoalstringThe main objective the service should optimize for.
serviceConfig.knowledgeScopestring[]Topics, policies, or domain boundaries the avatar should focus on.
serviceConfig.capabilitiesstring[]Extra capability statements appended to the selected base service.
serviceConfig.escalationTargetstringWhere the avatar should route users when a human handoff is needed.
serviceConfig.welcomeMessagestringPreferred greeting or first-reply behavior for this service instance.
serviceConfig.placeholderstringService-specific input placeholder shown in the UI.
serviceConfig.promptAddendumstringExtra host instructions appended after the base service rules.
We can provide integration service for your team: iframe setup, secure host messaging, custom service definitions, role prompts, avatar selection, UI branding, and deployment support. This is designed to be easy to integrate, and if you want, our engineers can do the implementation for you.