Contact us

Rebel AI Studio by Codemate

The Agentic UI platform

Generate the right interface at the right moment — govern, operate, and monitor what your users see.

Built for the apps you already manage — Codemate's team handles initial integration, rules, and AI training.

Request live demo

Modern applications should adapt to the user — not the other way around.

Rebel AI Studio adds an intent layer to your app — chat, smart search, voice, or contextual triggers — that turns a user's request or need into the right interface, on the spot. It runs alongside your existing UI, drawing on your design system components and your business rules.

Your existing UI can stay as it is. The intent layer coexists with it — same components, same rules, but a second way in for users who'd rather describe what they need than navigate to it.

Three industry terms — Agentic UI, GenUI, A2UI — describe what Rebel AI Studio is built on. The platform implements all three, and adds the management and operations layer that makes them shippable.

Example: Traditional chat UI vs. GenUI chat Customer Service AI-Assistant Online Do I have any unpaid bills? You can find your Bills under Billing → Invoices. A link would have been helpful… 😟 Message… Do I have any unpaid bills? A link would have helped… 😟 Answer guides the user — but requires navigation and extra steps. Customer Service AI-Agent Online Do I have any unpaid bills? UNPAID INVOICE €50.00 Due yesterday Pay now Switch to an e-invoice? Yes No Message… Do I have any unpaid bills? Intent is translated into a task-specific interface — ready for action.
  1. Agentic UI Paradigm

    Autonomous agents proactively drive the user experience — combining text replies with real, interactive UI components.

    The platform that lets your business ship and operate Agentic UI safely — with allow-lists, IdP-bound permissions, and audit trails.

  2. GenUI Capability

    LLMs dynamically generate UI elements at runtime instead of selecting from static, hardcoded layouts.

    Assembles the right interface for each user intent — from your existing design system components, in real time.

  3. A2UI Protocol

    Google's open-source declarative JSON spec for streaming UI from agents to native clients. Cross-platform, no arbitrary code execution.

    A2UI-aligned — renders across Flutter, React, Vue, and Web Components, with a component catalog you control.

The intent layer is just the start — the platform lets you monitor the overall system usage and helps you evolve it

Every interaction can be analyzed: what users actually tried to do, what they wanted that isn't there yet, what terms they used for it, and how they felt along the way. Topics surface automatically across sessions; sentiment is tracked at the start and end of each conversation; your team can query the data in plain language. Static UIs can't reveal any of this — only an intent layer can. Your team turns those signals into change: refine prompts, tune agent behavior, version, roll back, ship the next iteration — from one management console, no engineering tickets needed.

Already in production with several enterprise customers — here's one. More case studies landing soon.

Moi ApulAInen — the conversational AI assistant in the Mun Moi app
Moi Mobiili

How Moi Mobiili brought AI into the customer experience

Codemate built the Moi AI assistant for Moi Mobiili — a conversational AI layer integrated into the Mun Moi app so customers can manage subscriptions and billing simply by having a conversation.

Read the case study

At a glance

Rebel AI Studio is an Agentic UI platform: a runtime that generates the right interface from your own components, governance bound to your rules, and a management console where your team edits prompts, tests changes, and measures impact.

Capabilities

Works with Flutter, React, Vue, and Web Components

Compatible with Gemini, OpenAI, and other leading LLMs

Hosted by us, or self-hosted in your own infrastructure for enterprise; authenticates via your IdP

Business owners edit AI behavior — no coding or engineering tickets needed

Versioning, rollback, and per-interaction audit trail

Native data-warehouse export (Google BigQuery)

How it works

  1. 01

    Intent

    Typed, spoken, or contextual

  2. 02

    AI reasoning

    Agent interprets the request

  3. 03

    Policy checks

    Allow-lists and role permissions

  4. 04

    Component catalog

    Approved design-system components

  5. 05

    Rendered UI

    Real interface, not just text

  6. 06

    Analytics

    Tagged with prompt version

Management console

Where AI behavior is edited, tested, versioned, and audited — without engineering tickets per change.

Edit AI behavior
Business owners change prompts, FAQ entries, and guardrails — no code required.
Test before you publish
Side-by-side prompt comparison and blind tests; staged rollouts.
Versioning + audit
Per-interaction audit trail; roll back to any prior working state.
Access control
Permissions bind to your IdP; scoped edit rights per role.
Analytics + export
Natural-language reports linked back to original sessions; native BigQuery export.

Why teams choose Rebel AI Studio

Speed up delivery

Reuse your existing components instead of building new UI surfaces. Once Rebel is integrated and tailored to your stack, prompt changes and new iterations ship without a frontend rebuild.

Govern every action

Permissions bind to your IdP, components are allow-listed, and every action is attributable. The AI works inside your rules — not around them.

Keep things simple

Show only what is relevant right now. Rebel assembles the right interface for each moment instead of giving everyone the same crowded screen.

Insight-driven iteration

Built-in analytics reveal what users were trying to do and where their path broke. Natural-language reports and BigQuery export turn signals into next steps.

Adapt to different users

Real-time interface adjustments based on user intent, role, and context — without per-segment UI projects.

Trace every decision

Every recorded interaction is tagged with the exact prompt version active during it, so audit and rollback are always possible — for AI agents and human edits alike.

Where it lands

Whether the user is your customer or your colleague, the pattern holds: someone describes what they need in plain language; the AI pulls the right context, assembles the right interface, and writes clean structured data back to your systems.

Who it's for: Service ownersProduct teamsEngineering teamsOrganizations
  1. 01

    Augmented customer support

    The AI works alongside the support agent, not instead of them. It pulls case history, suggests next actions, and drafts structured replies for the agent to approve. Faster handling, fewer training cycles, the human stays in the loop.

  2. 02

    Customer self-service inside your app

    A consumer types "book the next paid ride to the airport" or "where's my refund?" The AI assembles the right screen, calls the right APIs, and either completes the transaction or escalates cleanly — through your existing systems, with your auth.

  3. 03

    Sales transactions, recommendations, and comparisons

    Natural-language product search and comparison built into the same app where customers already buy. "Show me the ones with X under Y euros, including delivery." The AI surfaces the right options from your real catalog and pricing data, with sources cited back to the original records.

  4. 04

    Field service and maintenance

    A technician on site dictates what they see; the AI pulls the relevant manuals, the equipment's service history, and the customer's prior tickets, then captures the field report back as structured data the maintenance CRM expects. The notebook-then-retype workflow disappears.

  5. 05

    Sales in the field

    A rep talks through a meeting; the AI captures structured notes, surfaces relevant comparison data — competitor pricing, related products, account history — and writes a clean activity record into the CRM. The rep's job is the conversation, not the data entry afterwards.

  6. 06

    Internal workflows (HR, finance, procurement, ops)

    Any internal team that runs repetitive structured work. An employee describes what they need; the system assembles the right form, applies the right approval routing, and submits the request through the systems you already have. The same pattern as customer self-service, pointed inward.

07

Any input → structured data

The thread that runs through all of the above. Any free-form input — voice notes, photos, scanned documents, plain-text descriptions — becomes the clean structured data your downstream systems expect. Useful anywhere the workflow today involves a person retyping their own input.

Request a live demo

This page sketches the shape of the platform. A live demo goes deeper — workflows, testing, analytics, agent orchestration — and gives us a chance to discuss how Rebel would fit your stack.

Request: Rebel AI Studio live demo

FAQ

What is GenUI (Generative UI)?

GenUI is the technical capability of an LLM dynamically generating UI elements at runtime, instead of selecting from a static, hardcoded layout. Rebel uses GenUI to assemble the right interface for each user intent — drawing components from your existing design system, governed by your rules.

What is Agentic UI?

Agentic UI is the design paradigm where autonomous agents proactively drive the user experience — combining text replies with real, interactive UI components. Rebel is an Agentic UI platform: the agent-driven UX layer plus the management console your business uses to govern and operate it.

What is A2UI (agent-to-UI)?

A2UI is Google's open-source declarative JSON protocol for streaming UI definitions from an agent to a native client. It guarantees security and cross-platform native rendering without executing arbitrary code. Rebel is A2UI-aligned: agents emit intent, components from your allow-listed catalog render across Flutter, React, Vue, and Web Components.

Who edits the AI — engineers or business owners?

Business owners edit prompts, FAQ entries, and conversational guardrails through the management console — no code required. Engineering owns technical components and integrations. Role-based permissions enforce the split, and every change is attributable to a real person in your identity provider.

How do you safely change a prompt in production?

Edit and test in an environment separate from production, run new prompt versions side-by-side against the same questions (blind comparisons supported), then deploy to a limited group before rolling out widely. Every change is versioned and reversible at any time.

Can we audit what the AI actually saw and said?

Yes. Every recorded interaction is tagged with the exact prompt version that was active during it, so post-hoc reviews always reflect the AI's real state at that moment — not what's live now.

Does Rebel AI Studio replace our frontend?

No. It uses your existing design system and components. Think of it as a layer that decides which components to show, when, and why.

Can we control what the AI can do?

Yes. Guardrails, policies, and audit trails ensure safe behaviour. Every action and component is allow-listed by your team.

Which AI models are supported?

Google Gemini, OpenAI, and other leading LLMs. The platform is model-agnostic.

How do we measure impact?

Built-in analytics aggregate real end-user interactions across the deployed application. The console surfaces what users were trying to do, which components were used, and where their path broke. Reports identify recurring themes and suggest next steps, and each finding links back to the original sessions — verifiable in the source data, not an AI summary. Native BigQuery export keeps unlimited history for deeper analysis.

Can this control AI agents too?

Yes — that's a core use of the platform. Rebel manages and monitors AI agents from the same management console, with full audit, versioning, and rollback for every agent decision.

Do I need to use AI for everything?

No. AI interprets intent, but the components themselves remain traditional, well-tested code from your design system.