The Problem

Every user wants something different. Your sales lead needs a pipeline view. Your CFO wants financial KPIs. Your account manager needs a client overview. And every new customer, every new role, every new use case means another dashboard variant your developers need to build and maintain.

Static templates don’t scale. Your engineering team spends months building views that are outdated by the time they ship — and the next user wants something slightly different anyway.

Meanwhile, your marketing team runs campaigns across channels and regions, but every visitor lands on the same generic page. Personalisation would convert better, but maintaining dozens of landing page variants per campaign is unsustainable.

And your AI agents? They communicate through text. But text is a poor medium for showing data, comparing options, or guiding decisions. Your agents need a canvas, not just a chat window.

What We Build

We build generative UI — interfaces that create themselves based on context. Not templates. Not pre-built variants. Interfaces that generate in real time for each user, each request, each situation.

This is one of the next big shifts in AI: instead of bolting a chatbot onto an existing product, the entire interface adapts. Every view is unique because it’s generated, not selected from a menu.

What this means in practice:

• Dynamic dashboards — users describe what they want to see, and the interface assembles: charts, tables, KPIs, filters — composed from your data in seconds, not sprints
• Hyper-personalised landing pages — every visitor gets a page generated for them: their language, their industry, their pain points, their referral source — not one generic page for everyone
• Agent interfaces — give your AI agents a visual canvas to communicate with users, show structured data, render interactive components, and guide decisions — beyond plain text
• Adaptive reports — each client gets a report that looks like it was designed for them, because it was — generated from their data and preferences

Why This Matters Now

Most companies adding AI to their product are bolting on a chatbot. That’s table stakes. The companies that will stand out are the ones whose entire interface adapts to their users.

Generative UI is the difference between “we have AI” and “our product feels like it was built for each user.”

What’s changing:

• AI agents like OpenClaw and Hermes communicate through text — but text is limiting. Generative UI gives agents a canvas to show data, render components, and interact visually
• Landing pages are becoming indistinguishable — every competitor uses the same templates. Hyper-personalised, generated pages are how you stand out
• Dashboard fatigue is real — your users don’t want another static view. They want to describe what they need and get it instantly

How It Works

Your data, your brand, your rules

We don’t replace your product. We add a generative layer on top of your existing data and APIs. Your design system, your branding, your data governance rules — all respected. The AI handles layout and composition. You control what’s possible.

Not random — intelligent

Generated doesn’t mean unpredictable. The AI works within constraints you define: which data sources are available, which components can be used, what layouts are appropriate. Think of it as a very fast, very flexible design system that assembles itself.

Built to ship, not to demo

We build production systems, not prototypes that break under real usage. That means proper error handling, performance optimisation, and fallbacks when the AI’s output needs guardrails.

See It In Action

We built Cuttlekit — our open generative UI toolkit — to prove this approach works. You describe what you want. An LLM generates a working interface. No code, no templates, no rigid layouts.

Good Fit If…

• You’re a SaaS company whose users keep asking for custom views and dashboards
• You serve multiple user roles or industries from a single product
• Your engineering team spends too much time on UI variants instead of core product
• You run multi-market campaigns and need localised landing pages without managing dozens of variants
• You’re building AI agents that need more than text to communicate effectively
• You deliver recurring reports to clients and the customisation is eating your team’s time