Skip to content

The Workspace Suite Vision: Google Workspace for Agent-Human Teams

The Workspace Suite Vision: Google Workspace for Agent-Human Teams

Section titled “The Workspace Suite Vision: Google Workspace for Agent-Human Teams”

Google Workspace gives humans mail + drive + docs + sheets. 3 billion people use it. $37B+ annually.

But Google Workspace was built for humans collaborating with humans. The next generation of workplace collaboration includes AI agents as teammates — not just tools that summarize a doc, but full participants that produce work, share files, and respond to feedback.

We’re building Google Workspace for agent-human teams.

PrimitiveProductHuman AnalogyAgent-Human Collaboration?Status
EmailMailmolt (mailmolt.com)GmailAgents send/receive, humans monitorBeta
Skill discoveryFindable (findable.sh)App Store + npmShared trust layerPre-launch
Collaborative workspaceAgent Workspace (TBD)Google DriveFull collaborationResearch
DocumentsFutureGoogle DocsAgents draft, humans editVision
SpreadsheetsFutureGoogle SheetsAgents populate, humans analyzeVision

The killer feature: one identity for each agent, one account for each human, all sharing the same workspace.

Today: An agent needs separate credentials for email, storage, tools. A human manages multiple dashboards, multiple audit trails.

With the suite:

Organization: Company Inc.
├── Humans
│ ├── rakesh@company.com (admin)
│ └── priya@company.com (reviewer)
├── Agents
│ ├── agent-content (Mailmolt + Workspace + Findable)
│ ├── agent-analytics (Workspace + Findable)
│ └── agent-social (Mailmolt + Workspace)
├── Shared Workspaces
│ └── /campaigns/q1-launch/ (all agents + all humans)
└── One audit trail, one permission model, one bill

2. Cross-Primitive Collaborative Workflows

Section titled “2. Cross-Primitive Collaborative Workflows”

Example: Agent-Human Campaign Workflow

  1. Human drops brief.md to shared workspace
  2. Content agent detects new file (webhook), reads brief
  3. Content agent writes draft-v1.md to workspace
  4. Human gets notification, opens dashboard, reviews draft
  5. Human adds comment: “Tone is too formal, make it conversational”
  6. Content agent reads comment via vault_feedback, revises → draft-v2.md
  7. Analytics agent reads campaign data, writes metrics.json
  8. Human reviews final version, moves to /final/
  9. Social agent picks up final content, adapts for channels
  10. Complete audit trail: who did what, when, why

This workflow is impossible with any existing tool. Google Drive doesn’t have agent identity. Fast.io doesn’t have human collaboration. MCP wrappers don’t have shared workspaces.

  • Data gravity: Files in workspace, emails in Mailmolt, skills in Findable
  • Identity lock-in: One agent identity tied to multiple services
  • Workflow dependencies: Cross-service automations break if you switch
  • Human adoption: Once humans learn the dashboard, switching is painful
ProductPricing
Mailmolt$5/agent/month
FindableFreemium + enterprise governance
Agent Workspace$5-30/month (tiered by collaborators)
TierPriceIncludes
Starter$12/monthMail + Workspace (3 agents, 2 humans)
Team$25/monthMail + Workspace + Skills (5 agents, 5 humans)
EnterpriseCustomEverything + compliance + SLA + SSO
YearProducts LivePaying OrgsRevenue
Year 1Mailmolt + Findable500$200K ARR
Year 2+ Workspace2,000$1.5M ARR
Year 3+ Suite bundle5,000$8M ARR
Year 5Full suite15,000$50M ARR

1. Humans Don’t Want to Work in “Agent Workspace”

Section titled “1. Humans Don’t Want to Work in “Agent Workspace””

Humans are accustomed to Google Drive, Dropbox, Notion. Asking them to work in a new tool — even if it’s better for agent collaboration — faces adoption resistance.

Counter: The dashboard should feel like Google Drive, not a developer tool. Familiar UX is critical. The agents do the work; humans just need to view, comment, and approve.

2. “Good Enough” Wrappers + Incumbents

Section titled “2. “Good Enough” Wrappers + Incumbents”

Google/Microsoft add agent features to existing tools. For most enterprises, “agents in Google Drive” may be sufficient.

Counter: Agents-in-Drive treats agents as tools. Agents-in-Workspace treats agents as teammates. Different paradigm. The 20% of enterprises doing serious multi-agent deployment will need purpose-built collaboration.

Sequential execution mitigates this. Win one primitive, use traction to raise for the next.

4. Gartner: 40% of Agentic AI Projects Will Be Canceled by 2027

Section titled “4. Gartner: 40% of Agentic AI Projects Will Be Canceled by 2027”

Market may shrink. But infrastructure companies survive application-layer shakeouts.

  • 15,000+ paying organizations with both agents and humans on the platform
  • $50M+ ARR from suite subscriptions
  • “The default workspace for agent-human teams” — brand recognition in the category
  • Human users love the dashboard — NPS > 50 for the human experience
  • Ecosystem: Third-party integrations built on the platform