Enterprise AI workspaceprogress visible from day one

Employees work in the same space — projects, files, conversations and every execution's usage stay on the company side.

Daily use

Ready to use

Workbench

From first sign-in, a clear next step

The workbench guides employees to request a workspace. Once an administrator approves it, they can create projects and start working.

The web app and desktop client offer the same capabilities — projects, files, conversations and tasks stay in sync. Tasks run in the environment the company selects, supporting SaaS, private deployment and appliance.

  • Projects, workspaces and remaining quota in one view
  • Switch to expert view when more control is needed
Workbench: continue where you left off, goal input and projects from the last 24 hours (demo data)

Workspace

Project data and context stay in the workspace

Employees work inside approved workspaces. Files, conversations, tools and execution records are kept together and can be shared or handed over by permission.

  • Project context persists across devices and sessions
  • Multi-person collaboration, sharing and handover
Project session in a workspace: projects and capabilities on the left, conversation in the middle, project info and token usage on the right (demo data)

Apps

Notes and task system already in the workspace

The workspace comes with a notes system and a task system: jot down project notes, track what needs doing — ready to use. Need other services? Connect them over MCP and call them from the same workspace.

  • Notes system and task system included
  • Connect other services over MCP
Notes and task system open in a workspace: note directory and task list (demo data)

Browser

AI opens its own browser, people can take over at any time

The operation is visible in real-time. Employees can take over at any time, and colleagues can watch too — easy for review and handover.

  • Execution visible in real-time, take over anytime
  • Colleagues can watch along for review
Browser panel on the right side of a workspace: AI operating a web page, ready for human takeover (demo data)

AI quota

Every AI usage traced to an owner

Employees see their balance and current plan; administrators review usage by workspace, project, runtime and model.

  • Employees check remaining quota anytime
  • Usage verified by workspace, project, runtime and model
Workspace interface showing project token usage (demo data)

Team collaboration, managed for your team

Clear visibility, manageable control, stable operation. People, information, processes and assets in one ecosystem.

Collaboration

One project one group, members follow project permissions

Project members are automatically added to the project group. When they leave the project, they no longer see subsequent discussions. Supports private chat, @mentions, unread and read status.

  • Members, permissions and sharing managed inside the project
  • Leaving the project stops further discussion visibility
Project group chat: discussions and members in the same project (demo data)

Share center

Completed work reusable in the next project

Templates, processes and project data can be shared by role and expiry, ready for the next project. Permissions can be revoked at any time.

  • Share by role and expiry
  • Permission revocation centralised
Share centre: projects and sessions by role and expiry (demo data)

Agent market

Pick an agent by task, install and use

Employees only see agents the company has opened. Publishing, upgrading and removal are decided by the administrator.

  • Choose by task, skills and processes already configured
  • Platform maintains updates, employees don't manage versions
Agent market: agents grouped by task with install entry (demo data)

Runtime

Independent runtime, files and tools stay put

Each workspace has independent compute and storage. Employees can run code, save files or open services. Administrators review resource usage.

  • CPU, memory and disk allocated per workspace
  • Running services openable by approval
Workspace instance page: resource scheduling overview compared with allocated quota (demo data)

Admin console

Daily administration in one console

Administrators handle workspaces, member permissions, model access, approval requests and login logs here.

  • Who can use what, decided by role
  • Model providers and keys managed centrally
Admin console home: workspaces, agents, model access and task overview (demo data)

Work anywhere, on any device

Desktop and web client

Desktop client workspace overview (demo data)
Desktop client project conversation and file panel (demo data)
Discuss a pilot