Skip to content

The platform

Connect your data. Build the application. Put it to work.

Three pillars, one governed workspace. Dimension pulls the systems you already run into a live data layer, builds applications on it with AI Composer and lets them act on your estate, with a person approving every change.

Start free trialGet in touch
  1. Data

    Every application starts with the right data

    Connects the systems you already run, through their APIs, into one live, consolidated data layer that every application queries directly. Read-only, so building and reporting never touch production.

    • Any API

      Cloud platforms, identity providers, security tools, HR and business systems, in one consolidated layer. No CSV exports or stale mirrors.

    • Describe the source

      AI Composer configures the connection and schedule for you.

    • Read-only by design

      Building and reporting never touch production.

  2. Studio

    Where security applications take shape

    Where applications are built. Describe what you need and AI Composer builds it: data connections, logic and a working interface, on your real data and behind your own authentication.

    Or build with your own AI coding agent over MCP. There's no backend to manage and no development team needed. Build in a sandbox, then promote to production.

    • Run, customize or build

      Run a ready-made application from the App Library, reshape one that's close in plain language or describe the one that doesn't exist yet.

    • A sandbox on real data

      Build against your live data layer, behind your real authentication and permissions.

    • You choose who sees it

      Keep an application private, share it with your team or open it to everyone.

  3. Actions

    Where your applications get to work

    Where applications change live systems: disable a user, open a ticket, update a cloud configuration, trigger a device policy. Every action is approved by a person and written to the audit log.

    An application can only call the actions connected to it, and your existing permissions stay in place.

    • Approval on every action

      A person signs off before anything changes a live system.

    • Least privilege by default

      Applications see only what you connect to them. Actions run within the access people already have. Nothing is elevated.

    • A record of everything

      Who asked, who approved, what changed and when, in an audit log that can't be edited and can be exported for auditors.

Works where you work

Build over MCP, from the tools you already use.

Build, edit and publish applications from Claude Code, Claude Desktop, VS Code or OpenAI Codex over MCP. It is the same enterprise foundation underneath: your live data, your permissions and an approval on every action.

Already built something in a sandbox? Open it in Dimension over MCP and it runs in production, behind your SSO, with audit and approvals in place.

Start free trial
Terminal

A Claude Code session: the Dimension MCP server is added, then Claude is asked to build an access review for finance from Entra ID and Workday. It checks the data feeds, finds Entra ID ready and Workday missing fields, then adds a Workday feed using credentials already in the Dimension Credential Vault. Then it builds the application in Studio, asks whether to publish it and, once confirmed, publishes it to app.dimension.io/finance-access-review. To connect it yourself, run: claude mcp add dimension https://mcp.dimension.io/mcp --transport http

Enterprise foundation

Building is easy now. Production is not.

Most homegrown applications stall before production, or ship without the controls they need. Dimension is that production environment, already built and maintained, so your team starts with the hard part done.

SSO and RBAC

Every application ships behind your SSO with roles and permissions, on every plan.

Audit log

Every view, change and action lands in a tamper-evident log your assessors can query or export.

Action approvals

A person signs off before anything changes a live system.

Hosting and tenancy

Dedicated, isolated tenancy on AWS, with regional residency.

Data at scale

Platform architecture built for estates of millions of assets, with backups and versioning maintained for you.

Safe by design

Live data stays read-only. Models are never trained on your data.

Start small. Scale when it's working.

Every plan runs on the same governed foundation: SSO, approval on every action and a full audit log from day one. Start on your own data with a 14-day free trial, no credit card and nothing to deploy. Move up when the team wants more applications, more sources or longer retention.

  • Starter

    $49

    per month, billed monthly

  • TeamMost popular

    $189

    per month, billed monthly

  • Business

    $1,500

    per month, billed monthly

  • Enterprise

    Custom

    volume, seats and rates

Every plan includes your first 10,000 records and SSO.

Compare plans

Questions

Technical questions

All questions

  • Any system with an API: cloud platforms, identity providers, security tools, HR and business systems. AI Composer configures the connection and sync schedule for you.

  • Only through Actions you explicitly connect. The data layer is read-only, so building and reporting never touch production. When an application needs to make a change, such as disabling a user or opening a ticket, a person approves it first, your existing permissions still apply and the action is written to the audit log.

  • AI Composer builds the application and the platform runs it. A running application can only read the feeds and call the actions connected to it. Your data is never used to train models.

  • In your own dedicated, isolated tenant on AWS, never a shared pool, with regional data residency.

  • Every view, change and action, in a tamper-evident log your assessors can query in the platform or export. Audit export and retention controls are included from the Team plan upward.

  • You build in a sandbox against your live data, behind your real authentication and permissions, then promote it to production when it’s ready. Every application is versioned, so you always know what changed and when.

  • Yes. Build, edit and publish applications from Claude Code, Claude Desktop, VS Code or OpenAI Codex over MCP. It’s the same enterprise foundation underneath, with the same data, permissions and approvals. Applications already built elsewhere can be imported and run in production.

  • SSO with roles and permissions is included on every plan. Business adds SCIM provisioning, and fine-grained RBAC is coming soon.

  • Nothing changes. Every application lives in a shared, versioned workspace with its own permissions, on a platform we host and maintain, so it doesn’t depend on one person.

Build it. Approve it. Run it.

Try Dimension free for 14 days on your own data. No credit card, nothing to install, and nothing reaches production without your sign-off.

Start free trialGet in touch