Products / AI Fabric Chat UI v0.3.0 Public Beta

Turn governed AI workflows into a real application experience.

Add retrieval evidence, clarification forms, confirmations, action results and backend-owned conversations to any web application—without moving identity, policy or business execution into the browser.

Explore live demo View on GitHub

Web Component
React
Spring Boot
Arabic & RTL
Apache 2.0

npm install github:Loom-AI-Labs/ai-fabric-chat-ui#v0.3.0

AI Fabric Chat UI embedded in an account support application

A presentation layer built around application control

Typed orchestration outcomes

Evidence, clarification, confirmation and structured results.

Backend-owned governance

Identity, tenant policy and execution remain in your application.

Four responsive layouts

Docked, floating, inline and full-screen.

Application-native extension

Branding, domain tools, action renderers and attachments.

Host application
AI Fabric Chat UI
Application API
AI Fabric Framework

The problem

Useful AI needs software boundaries, not only prompts.

A plain message list cannot explain retrieval evidence, ask structured clarification, collect confirmation or present action outcomes with enough context for a dependable product experience.

The product value

A reusable layer with an explicit host contract.

Chat UI supplies those reusable interaction surfaces while the host application keeps control of authentication, conversation persistence, policy and business operations.

Capability pillars

Five responsibilities, each visible at the integration boundary.

Conversation surfaces

Render backend-owned conversations in compact, full and embedded layouts.

Retrieval evidence

Present sources, documents and grounded answer context without raw payloads.

Structured clarification

Collect missing fields through explicit, application-friendly form contracts.

Action confirmation

Show proposed operations and capture deliberate user confirmation.

Host integration

Use Web Component or React bindings with Arabic and RTL support.

Ownership boundary

Product responsibility stops before application authority.

Chat UI owns

Rendering and interaction state for supported AI response contracts
Accessible message, evidence, clarification and confirmation components
Web Component and React integration surfaces
Responsive and bidirectional presentation behavior

Host application owns

Authentication, session issuance and conversation authorization
Backend requests, persistence and retry policy
Business action permissions and side effects
Product navigation, surrounding layout and customer support policy

Architecture

A narrow path through explicit contracts.

01 Host application
02 Chat UI surface
03 Backend conversation API
04 AI Fabric workflow

Quick start

Begin with one application-owned path.

Start from the public documentation, wire one bounded scenario and keep authorization and source data in the host.

Integration guide

<script type="module" src="/ai-fabric-chat-ui.js"></script>

<ai-fabric-chat
  endpoint="/api/assistant/query"
  locale="en"
  mode="embedded">
</ai-fabric-chat>

Live proof