A native assistant in the Designer

Describe the app you need and the assistant creates resources, fields, pages, permissions, and workflows for you. Keep prompting in natural language, or drop into the visual editor any time.

AI features inside your apps

Ship AI inside the apps you build. Ready-made function templates for document extraction, content generation, classification, and embeddings plug into your data as triggers. End users get AI-powered workflows, not just admins.

Guided design & audit workflows

Built-in assistant workflows for reviewing changes between versions, auditing role and permission matrices, scaffolding new features, and comparing configurations - grounded in your actual schema.

AI Assistant

From prompt to working app.

Describe what you need in plain English. The assistant produces a data model, pages, workflows, and permissions - a working first draft in minutes. Refine with follow-up prompts or jump into the visual editor.

  • Natural-language modelling of resources and fields
  • Page and form generation with preview
  • Workflow suggestions derived from your data model
  • Context-aware - knows your existing apps and conventions
xms.dinode.com/ai
xMS AI assistant
AI at runtime

Ship AI features inside the apps you build.

xMS comes with ready-made AI function templates. Plug them in as triggers on your resources - on insert, on update, on a schedule - and your apps get AI-powered behaviour without a separate ML project.

  • Document extraction — pull structured JSON from uploaded files
  • Content generation — draft emails, summaries, descriptions
  • Classification — categorise records as they come in
  • Embeddings — search records by meaning, not just keywords
  • AI consumption bundled in and metered separately - no surprise bills
xms.dinode.com/functions
Trigger: on insert · claim
→ ai_document_extraction
✓ 14 fields extracted in 2.1s

What AI inside your apps looks like in practice

Every one of these is built by attaching a ready-made AI function to a trigger on your resource. Configure the prompt, pick the model - the rest is wiring you already own.

Extract data from uploads

When a PDF, image, or Word doc lands on a record, have the assistant extract structured fields - supplier, date, amount, reference - into the record automatically.

Classify on arrival

New claim, case, or request comes in? Categorise it by type, severity, or department as the record is saved, and route it to the right queue with no manual triage.

Draft the first response

Generate a first-draft email, acknowledgement, or response letter against your templates. A human reviews and sends - minutes instead of hours.

Summarise long records

Long case, claim, or application history? Keep a running AI-generated summary field up-to-date so a reviewer sees the gist in one line instead of scrolling.

Semantic search

Embed records and documents on write. Users find related items by meaning - "policies like this one" or "prior cases with similar facts" - not just keywords.

First-pass assessment

Against your eligibility rules or scoring criteria, draft an assessment for each incoming item. A human accepts, rejects, or refines - but the AI carries the boring first pass.

Prefer to build from Claude Desktop or Cursor? The xMS Designer exposes an MCP server too. Point any MCP-compatible client at it and you can design apps, edit permissions, and compare versions from your AI client of choice - it speaks to the same tools the Designer's built-in assistant uses.

See AI-native xMS in action.

Book a demo and we'll walk you through the assistant, MCP, and an in-app AI feature - live.