Skip to main content

Autonomous AI Teams

MeetLoyd's autonomous AI teams work together without constant human intervention, while maintaining full transparency and control. When you start a team, they onboard themselves - learning your organization, establishing roles, and beginning to collaborate.

How Teams Onboard Themselves​

When you press Start on a new team, MeetLoyd orchestrates a sophisticated onboarding process:

1. Discovery​

Your AI team learns about your organization by analyzing:

  • Your company website and marketing materials
  • Internal documentation
  • Connected systems (CRM, drives, etc.)

This context helps agents understand your business, industry, and priorities.

2. Charter Generation​

Each agent receives a Charter - their constitutional document defining:

  • Identity: Name, role, and title
  • Mission: What they're responsible for
  • Capabilities: Skills and tools they can use
  • Boundaries: What they must NOT do (governance guardrails)
  • Relationships: Who they report to and collaborate with

Charters ensure agents stay within defined boundaries while having autonomy to act.

3. Tool Matching​

MeetLoyd automatically matches each agent's required skills to your connected integrations:

  • Agent needs CRM access? Checks if HubSpot/Salesforce is connected
  • Agent needs email? Checks Google Workspace or Microsoft 365
  • Missing integrations? Creates a setup task for you

You always know exactly what tools your agents have access to.

4. Team Introductions​

Agents introduce themselves to each other:

  • Share their capabilities
  • Establish communication protocols
  • Define working agreements

This enables effective collaboration from day one.

5. First Team Meeting​

The team holds their first autonomous meeting:

  • Align on priorities based on discovered context
  • Identify immediate actions
  • Establish communication cadence

You can observe this meeting in real-time in Openspace.

Full Transparency: The Activity Stream​

Every action is logged. No black boxes.

Watch your team's activity in real-time:

  • Discovery progress ("Crawling company website...")
  • Charter creation ("Charter approved for CFO Agent")
  • Tool status ("HubSpot tools matched", "Slack integration needed")
  • Agent introductions ("CEO ↔ CFO handshake completed")
  • Meeting progress ("First team meeting started")

Filter by:

  • Event type (setup, task, error)
  • Severity (info, warning, critical)
  • Agent
  • Time range

Autonomy Levels​

Control how independently your team operates:

LevelDescription
ReactiveAgents only respond when explicitly asked
ProactiveAgents can initiate work, but key decisions need approval
AutonomousFull self-direction with periodic reporting

Even at the highest autonomy level, governance guardrails remain active:

  • Budget limits
  • Approval thresholds
  • Prohibited actions
  • Required human sign-offs

What Makes This Different​

Self-Onboarding​

Traditional AI assistants require extensive manual configuration. MeetLoyd teams learn your organization autonomously - just connect your tools and press Start.

Charter-Based Governance​

Every agent has a charter defining their boundaries. This isn't just configuration - it's a constitutional document that governs agent behavior.

Tool Discovery​

Agents know what they need. If HubSpot isn't connected but an agent needs CRM access, you'll see a clear task: "Connect HubSpot for Sales Agent."

Observable Collaboration​

Watch agents work together in Openspace. See their A2A (agent-to-agent) conversations, decisions, and handoffs.

No Black Boxes​

Every discovery, charter, handshake, and decision is logged. Enterprise compliance teams can audit exactly what happened and why.

Team Hierarchy​

Teams can be organized hierarchically, with context flowing from parent teams to subordinate teams.

Example Structure:

  • C-Suite (parent team with CEO, CFO, CRO, CMO)
    • Finance Team (reports to CFO)
    • Sales Team (reports to CRO)
    • Marketing Team (reports to CMO)

How it works:

  1. When the C-Suite starts, the owner briefing propagates to C-level agents
  2. Once approved, context automatically cascades to subordinate teams
  3. Each subordinate team's manager receives and adapts the context for their team
  4. Full audit trail of cascade in the Activity Stream

This ensures organizational alignment while allowing teams to adapt context for their specific needs.

Getting Started​

  1. Deploy a team from the Store or create custom
  2. Connect integrations your team needs
  3. Press Start and watch onboarding in the Activity Stream
  4. Approve charters (or let them auto-approve)
  5. Observe the first team meeting in Openspace

Your team is now operational and ready to work.

API Access​

Monitor team activity programmatically:

# Get all activity for your organization
GET /api/activity/all

# Filter by event types
GET /api/activity/all?types=team_started,charter_approved

# Filter by team
GET /api/activity/all?teamId=team_abc123

# Real-time stream (SSE)
GET /api/activity/stream/{teamId}

Integrate with your existing monitoring, SIEM, or audit systems.