A digital human and conversational AI platform.
IAMX lets you create an AI avatar or digital worker and choose how people meet it: a website assistant, Unreal MetaHuman or on-site kiosk. Manage voice, persona, knowledge and permitted tools, then prepare the rendering and interaction for each destination.
Choose the job and the presentation separately
A digital human is the interface; its role defines what it helps someone do. Build a digital worker for employee questions, reception or product guidance, a customer service representative for approved support knowledge, or an AI chatbot for text and voice conversations. The same authoring tools can prepare AI tutors, employee training assistants, roleplay NPCs and live virtual presenters.
Give each role its own instructions, relevant documents and permitted tools. An order lookup, booking request or CRM update needs a separately configured service; selecting a role does not install those integrations. Choose a GLB avatar, image-based character or Unreal scene according to where the conversation takes place.
At a glance
| Area | Capability and scope |
|---|---|
| Character configuration | AI persona, instructions, voice, knowledge, conversation memory and permitted tools are managed in the web panel. |
| GLB avatars | Browser-rendered 3D models. Body animation requires a compatible skeleton; facial movement requires supported morph targets. |
| Unreal MetaHumans | An Unreal plugin connects the character to IAMX. Browser delivery uses a streamed Unreal scene. |
| LiveFace | An alternative image-based avatar path, with its own generation and rendering requirements. |
| Conversation | AI chat and voice assistants with provider-dependent multilingual support. Voice access requires browser permission. |
| Actions | Inactivity rules and explicit model action codes can trigger configured animation and media. |
| Website templates | Eight presentation templates. The character-only transparent template applies to GLB rendering. |
| Lip sync | Live speech uses the conversation pipeline. Uploaded action narration can reuse prepared speech-to-face animation data. |
| Knowledge and tools | RAG knowledge assistants and API-connected AI agents, subject to plan access, enabled sources and integration setup. |
| Rendering cost | GLB uses the visitor device. Streamed Unreal needs local or hosted rendering hardware. |
AI avatar platform FAQs
Is IAMX an offline video generator?
The main workflow is live interaction: a visitor speaks or types and the character responds. Prepared videos and narration can also be used as action media.
Does every model support facial animation?
No. The model must contain the facial controls expected by the selected mapping. A static mesh cannot acquire facial expressions just by uploading it.
Can an avatar complete business actions?
It can invoke configured actions and integrations. The connected system, permissions and application logic determine what the action actually does.
Is performance guaranteed?
Response time depends on the provider, audio processing, rendering, device and network. Evaluate the configured experience under representative conditions.
Documentation and ownership
IAMX is operated by IAMX LIVE DIGITAL HUMAN BİLGİ TEKNOLOJİLERİ VE YÖNETİM DANIŞMANLIĞI A.Ş. Product descriptions on this page are maintained by IAMX. Illustrations and demonstrations elsewhere on the site should not be read as measured customer outcomes.
Choose the destination around the visitor's task
| Destination | Typical configuration | Start here |
|---|---|---|
| Website avatar | A compact character beside a page, with voice or text input and enabled product or service knowledge. | Widget layouts and publishing |
| Unreal experience | A character in your Unreal scene, connected to IAMX conversation and the Blueprint actions you configure. | MetaHuman integration |
| Physical kiosk | A dedicated display and microphone, with a visitor task and operational support suited to the location. | Kiosk planning and evaluation |
| Character authoring | A modular, ready-made or uploaded GLB avatar, with scene, motion, authored media and conversation testing. | Avatar Studio workflow |
The conversation configuration and the renderer are separate choices. A character's knowledge does not create a 3D rig, and choosing a photoreal presentation does not provision a streaming host.
From source material to a working conversation
- Define the character's role
Write what it should help with, how it should speak and what to do when it does not know an answer. Select the conversation model and voice appropriate for the languages and response style you need.
- Grant specific knowledge and tools
Upload approved documents to the workspace library and enable the relevant files for the character. Define external tools separately and attach only the ones it should use. Uploading a workspace document or defining a tool does not automatically enable it for every character.
- Prepare the presentation
For GLB, test the model's facial shapes and body animations in Studio. For Unreal, configure the plugin and animation flow in the target project. Check the final camera, sound and interaction on the intended device.
- Test, publish and maintain
Use representative questions, missing information and service failures in a pilot. Publish the saved scene where applicable, review the visitor-facing result and repeat relevant checks when documents, tools or providers change.
Compare the complete cost of the experience
IAMX plan and usage
The plan sets character capacity, knowledge storage, feature access and conversation quota. Quota units depend on the model and speech configuration; they are not a fixed number of users, spoken minutes or conversations.
Connected providers and rendering
Your own AI or speech provider account may incur separate charges. GLB rendering uses the visitor or kiosk device. Unreal streaming requires separately provisioned rendering capacity; public IAMX conversation plans do not include cloud-rendering minutes.
Use your intended model, voice, character and average exchange length in a pilot before estimating usage. Check the workspace's saved plan for its actual limits, especially when a custom agreement applies.
Build a character people can talk to.
Start with a model, give it a voice and knowledge, and test the complete experience before publishing.
Maintained by IAMX · Updated · Platform facts and requirements