# Product Platform Overview ## What is an agent control plane? Building agents is easy. Operating should be too. Empower your engineers to build reliable agents faster with a vendor-neutral control plane to deploy and run any agent in minutes. [Book demo](/content/talk-to-sales/index.html) [Start free trial](https://app.guild.ai/) Guild marketing graphic ## Your Stack - Control Plane - Live - MCP Gateway - CLI - AgentHub - AI Observability - AI Governance - Requests Routed Today 0 - Running Agents ## Agent usage is growing every week, but most teams still can’t answer the foundational questions. ### What agents are running? ### What can they access? ### Who owns them? ### What did they do at runtime ### How much do they cost? ### What happens if they fail? ## Run, monitor, control. All your agents in one place. Anyone can build agents. The control plane makes them useful. Accelerate your AI transformation with deployment, governance, and observability for every agent and model. ## Run all your agents in a single catalog Import any agent from any model or build new agents from scratch—all inside a unified vendor-neutral control plane that lets you connect the right tools for every use case. [Learn more](https://app.guild.ai/) Connect any tool and any model Credentials are managed at the org level, resolved at runtime, and never exposed to agent code. [Learn more](/content/product/mcp-gateway/index.html) ## Access, build, and fork in our CLI You shouldn’t have to log into a UI to control your agents. The Guild CLI to brings deployment, monitoring, and agent management directly into the environments your engineers already use. ## Make skills and agents discoverable Publish agents and skills to a shared catalog so the whole org can find and reuse them. Guild marketing graphic ### Claude ### OpenAI ### Gemini ### Grok ### DeepSeek ### Llama ### Mistral release-manager Checks changelogs, tags releases, and notifies the team before every deploy. Use agent ## Control your agents like production software Import any agent from any model or build new agents from scratch—all inside a unified vendor-neutral control plane that lets you connect the right tools for every use case. [Learn more](https://app.guild.ai/) ### Improve trust with AI observability You can’t improve what you can’t see. Guild makes it easy to monitor all your agents at once, with AI observability to identify and resolve issues, build audit trails, and improve trust over time. Guild marketing graphic #### Audit Log guildai Last 24 hours - Events 48,291 - Agents 47 - Sessions 3,842 - Tokens 128M All LLM Tools Auth Sessions - 14:32:03 Tool github_pulls_get code-reviewer - 14:32:02 Auth Credential resolved: github code-reviewer - 14:32:01 Session Session started code-reviewer - 14:34:00 Session Session started deploy-bot - 14:33:08 Session Session completed (8.1s) triage-agent - 14:33:07 Tools slack_post_message triage-agent - 14:33:06 Tool jira_transition_issue triage-agent - 14:33:04 LLM claude-sonnet-4 (6.2k tokens) triage-agent - 14:33:02 Tool jira_get_issue triage-agent Evals 12 runs - Determinism score -2332% +4pts vs last week - Retrieves order status 12/12 - Formats currency consistently 12/12 - Escalates refund request to human 9/12 3 of 12 runs diverged on this step -- pin the prompt template version. - Cites source doc in response 11/12 - Declines out-of-scope request 12/12 4 of 5 cases stable Last run 2m ago ### Govern access and create accountability Credentials are scoped with per-endpoint controls to limit access. LLM interactions are mediated, and human approval flows can be enforced as needed. Governance is built directly into the runtime. Guild marketing graphic #### GitHub - 4 allowed - 2 denied - GET getPullRequest - GET listCommits - POST createReview - POST mergePullRequest - DELETE deleteRepository - DELETE removeBranchProtection - GET listBranches - POST createIssue #### Slack - 4 allowed - 2 denied - POST postMessage - GET listChannels - POST setTopicChannel - GET getUserProfile - DELETE deleteMessage - POST kickFromChannel - GET getConversation - POST addReaction #### Jira - 4 allowed - 2 denied - GET getIssue - POST createIssue - POST addComment - GET searchIssues - DELETE deleteIssue - DELETE removeVersion - GET listProjects - POST transitionIssue #### Linear - 4 allowed - 2 denied - GET getIssue - GET listCycles - POST createIssue - POST assignLabel - DELETE archiveProject - DELETE removeTeamMember - GET listWorkflows - POST updatePriority ### Optimize token costs to control spend View spending by workspace, provider, model, or user to track ROI in real dollars, or apply rate limits to control usage and reduce spend over time. Guild marketing graphic #### Agent Spend This Month - Total $-761.41 - +12% from last month ↑ - PR Reviewer $0.00 - Issue Triage $0.00 - Standup Summary $0.00 - Deploy Notifier $0.00 - Doc Sync $0.00 #### Top 5 of 47 agents - 12.8M tokens #### Spend Limits Org-wide - Monthly Cap $-5669/ $500 - 84% used #### Platform Team Team - $0/$800 - PR Reviewer Agent $0/$150 Near limit - Issue Triage Agent $0/$180 Over limit - Deploy Notifier Agent $0/$100 - 5 limits configured 1 over · 1 near limit ## Enterprise grade security for your agents and beyond ### Security and data privacy - No training on your data - Domain verification, SAML, SSO, & SCIM - SOC 2 Type 1 compliance - ISO 27001 certified - GDPR compliant - Data encrypted at rest with AES 256 and in transit with TLS 1.2+ ### Advanced admin controls - Azure, Anthropic, Cohere, OpenAI, and self-hosted models - Knowledge verification settings - User-generated feedback on responses - Analytics dashboard for monitoring usage and responses - API access for building your own integrations ### AI expertise - Dedicated success team, priority support, and SLA - AI strategy consulting and integration guidance - Tailored onboarding and change management support ## Frequently asked questions ### What is an AI agent control plane? An AI agent control plane is the operational layer that runs above agent frameworks. It maintains an inventory of every agent, enforces credentials and permissions at the moment of action, logs every input and tool call, and provides reversibility. Frameworks like LangChain, CrewAI, or custom code build agents. The control plane governs them in production. ### What is the Guild platform? Guild is a vendor-neutral control plane for AI agents. It lets organizations deploy, govern, and observe every AI agent their teams run, regardless of the framework, model, or team that built it. Guild sits between the agents your engineers create and the systems those agents touch, unifying identity, credentials, permissions, and audit into one layer. ### Which AI agent frameworks does Guild support? Guild is framework-neutral. Deploy and govern agents built with LangChain, CrewAI, Mastra, custom code, or any other framework. This is Guild's BYOA (Bring Your Own Agent) architecture: agents come from wherever your teams already build them, and Guild provides the runtime and control layer that runs them safely in production. ### How is Guild different from an agent-building framework? Frameworks like LangChain and CrewAI help you build individual agents. Guild is the layer that runs and controls them at production scale: identity, credentials, permissions, observability, budgets, and audit trails, applied consistently to every agent across every framework and model. Guild does not replace frameworks; it operates the agents they produce. ### Who is the Guild platform for? Guild is built for organizations running AI agents in production: platform engineering teams standardizing how agents ship, security and IT teams enforcing governance across shadow AI, and engineering leaders scaling AI adoption without losing control. If your teams are deploying agents faster than you can govern them, Guild is the layer that closes that gap. ### Is the Guild platform enterprise-ready? Yes. Guild is SOC 2 Type 1 compliant, ISO 27001 certified, and GDPR compliant. Data is encrypted at rest with AES 256 and in transit with TLS 1.2+. The platform supports SAML, SSO, SCIM, and domain verification, and every agent action produces a full audit trail. See the Guild Trust Center for the full security posture. Deploy. Control. Transform with Guild. See Guild in action. [Book a demo](/content/talk-to-sales/index.html) [Start free trial](https://app.guild.ai/)