AI WikiAI Wiki
Relevance AI logo

Relevance AI

Free tier

The enterprise platform for AI agents your team can trust and manage at scale

Free tier available·All audiences·Powered by Multi-model (Anthropic Claude, OpenAI GPT, Google Gemini)·API available
Visit

Key strengths

No-code + MCP-based agent builder for both ops teams and engineersVendor-agnostic LLM support (Claude, GPT, Gemini, and more)Enterprise-grade governance: RBAC, audit logs, PII masking, SSO/SAMLBuilt-in Evals framework for pre-deployment and production quality checks1,000+ native app integrations with native CRM, email, and calendar triggers
Free tier + paid plans
Sydney, Australia
No ratings yet

Technical Setup & Integration

Build Methods

  • Invent (NL prompt): POST a plain-language description to the Invent endpoint; the platform scaffolds agents, tools, and eval suites automatically.
  • No-code builder: Drag-and-drop canvas for creating multi-agent workflows with conditional logic, tool calls, and trigger configuration.
  • MCP integration: Connect Relevance AI as an MCP server to Claude Code, OpenAI Codex, or any MCP-compatible copilot to build and manage agents programmatically.

Triggers & Orchestration

  • Native triggers: CRM webhooks (HubSpot, Salesforce), inbound email, calendar events, cron schedules, and custom-built triggers within the platform.
  • Multi-agent coordination: Agents can hand off tasks, escalate to humans, and operate in parallel pipelines (L3 Autopilot / L4 Self-Driving modes).

LLM Configuration

  • Agents are model-agnostic; swap between Claude Sonnet/Opus/Haiku, GPT-5/Mini, or Gemini Pro/Flash per agent.
  • Built-in Model Evaluation runs per-1,000-task benchmarks to recommend the best model by eval score per dollar.

Governance & Observability

  • RBAC: Granular permissions for build, edit, deploy, and agent-level access.
  • Audit logs: Every tool call and decision is logged and exportable.
  • OpenTelemetry (OTEL): Export telemetry to your observability stack.
  • Delta Share: Share data with downstream analytics platforms.
  • PII Masking: Automatic detection and redaction before data reaches the model.
  • SSO/SAML 2.0: Centralized identity management.
  • Multi-region data residency: Deploy in the region required by your compliance policy.