Meet Loyd
Loyd is your AI co-pilot inside MeetLoyd. Think of Loyd as a colleague who knows the entire platform inside-out and can help you get things done through plain conversation -- no menus, no documentation hunting.
What Can Loyd Do?
Instead of navigating settings screens, you just tell Loyd what you need:
| You say... | Loyd does... |
|---|---|
| "Set up a sales team that qualifies leads" | Creates a team from the right template, asks a few clarifying questions, and deploys it |
| "Make my support agent more formal" | Adjusts the agent's system prompt and tone |
| "Why isn't my SDR sending follow-ups?" | Checks agent status, connections, and configuration to diagnose the issue |
| "What's the difference between execution modes?" | Explains the concept in plain language |
Team Setup and Configuration
Loyd can create new teams from templates or from scratch, configure team behavior (execution mode, autonomy, collaboration), add or remove agents, and set up connections and integrations.
Agent Management
Loyd can create and configure individual agents, adjust instructions and personality, enable or disable tools, and configure handoff rules and escalation paths.
Troubleshooting
Loyd can monitor team and agent status, diagnose configuration issues, check connection health, and suggest fixes for common problems.
Learning and Support
Not sure how something works? Ask Loyd. It can explain any platform concept, walk you through configuration, or suggest best practices.
Where to Find Loyd
Loyd appears as a floating chat widget in the bottom-right corner of the dashboard. Click the sparkles icon to open the panel.
- Available on every page while you are logged in
- Conversation history persists across sessions
- Minimizes with a single click
Loyd's Watchdog Role
Loyd quietly monitors your AI teams in the background and alerts you when something looks off:
| Signal | What Loyd Does |
|---|---|
| Infinite loops | Detects agents stuck in repetitive patterns and pauses them |
| Cost spikes | Flags unexpected increases in API usage |
| Excessive provisioning | Warns when too many resources are being created |
| Unusual behavior | Surfaces patterns outside normal operating parameters |
When Loyd detects an issue, it can automatically pause the affected team and notify you.
Loyd in Openspace
When you interact with Loyd from the Openspace view (the business-user workspace), Loyd operates in L1 support mode -- optimized for answering questions and surfacing information without making administrative changes.
What Loyd can do in Openspace:
- Answer questions about your teams, agents, integrations, and workspace
- Look up status and configuration
- Create tasks for the admin team when something needs elevated privileges
- Remember your preferences across conversations
What requires admin action: Creating or modifying agents, teams, or tools; granting permissions; Google Workspace or Microsoft 365 admin operations; and running tasks directly. Loyd will explain what is needed and offer to create a task for your admin team so your request is not lost.
Workspace Awareness
Loyd automatically knows which workspace and team you have selected in the dashboard:
- Scoped results -- "Show me my agents" returns agents in your selected workspace, not everything
- Placement confirmation -- Before creating anything, Loyd confirms the target workspace
- Context switching -- Change the workspace selector and start a new chat to pick up the new context
- User identity -- Loyd sees your name and email so it can address you personally and route follow-ups correctly
If Loyd seems to be showing data from a different workspace, start a new conversation -- the workspace context is captured when the conversation is created.
Privacy and Security
- Conversations are stored server-side and tied to your account
- Loyd only sees data within your selected workspace context
- Loyd cannot access other tenants' data
- All actions are logged in audit trails
- Loyd's permissions are governed by the same authorization system as every other agent -- your CISO can view, revoke, or restrict Loyd's access
Managing Loyd
You can manage Loyd from Settings > Security in the dashboard:
| Setting | Description |
|---|---|
| Email Integration | Give Loyd a Google Workspace email so you can email him like a colleague, CC him on threads, and receive replies |
| Email Security | Control who can email Loyd: allowed domains, allowed senders, rate limiting, and strict mode |
| Erase Memory | Permanently delete all of Loyd's memories and start fresh (use with caution) |
Loyd's configuration (tools, system prompt, model) auto-syncs on server startup. No manual sync required.
Availability
| Plan | Loyd Capabilities |
|---|---|
| Starter | Full assistant with all tools in the dashboard |
| Growth | Same capabilities plus advanced team orchestration |
| Enterprise | Custom Loyd training on your processes |
In Openspace, Loyd always operates in L1 support mode (read-only plus task creation) regardless of plan tier.
Tips for Working with Loyd
- Be specific -- "Set up a sales team" is good, but "Set up a sales team for B2B SaaS targeting CTOs" is better
- Ask follow-up questions -- If Loyd's first response is not quite right, clarify what you need
- Use Loyd for learning -- Ask "why" questions to understand best practices
- Let Loyd guide you -- When unsure, describe your goal and let Loyd ask the right questions