Relevance AI is an enterprise-grade, no-code/low-code platform for building, recruiting, customizing, and scaling autonomous AI agents (multi-agent systems) that run business processes end-to-end. Core capabilities include an AI Tool Builder (drag-and-drop flows, flow control, and templates such as OCR, SQL, web scraping, Slack/Email integration, Airtable, Webflow), unlimited agents/tools across plans, Bulk Run for large-scale processing, knowledge base onboarding (file uploads, website scraping, third-party connectors), and agent templates/roles to accelerate pilots (e.g., sales agent, research, pricing analyst, quote automation). The platform supports multi-provider models (OpenAI, Anthropic, Cohere, PaLM, self-hosted options) and Bring-Your-Own-Keys (BYOK) on paid plans to reduce platform pass-through costs. Relevance AI advertises extensive integrations (1000–2000+ connectors). Pricing and billing use a credit-based model with two main credit types: Actions (counts of tool/flow executions; typically do not roll over) and Vendor Credits (cover LLM/third-party model costs; vendor credits roll over while subscription is active). Plans include a Free starter tier (limited credits), Pro and Team paid plans, and Enterprise/custom plans. All plans include core features (agents, tools, Invent, integrations); paid plans provide larger Action & Vendor Credit allowances and enterprise features. Top-ups for both credit types are available and billing is handled via credit card per docs. If customers supply their own third-party API keys on paid plans, Relevance AI may not charge Vendor Credits for those runs (confirm exact terms in billing docs/FAQ). Security and compliance: Relevance AI states SOC 2 Type II compliance and GDPR-aligned controls, encryption in transit (TLS) and at rest (AES-256), customer-managed keys available, data residency options (US, EU/UK, Australia), tenant isolation, RBAC/SSO/MFA, and enterprise features such as fine-grained access controls and VPC/private networking. The company states customer data remains the customer’s and that Relevance AI does not use customer data to train their models except in explicit partnerships; tools are described as ephemeral in some contexts while datasets remain on-platform (review retention policies per plan). Governance and observability include usage dashboards and counters (Actions & Vendor Credits) with alerts, per-agent and org-level usage tracking, logs and retention that vary by plan, prompt-injection mitigations, human-in-the-loop workflows, and governance agents. Enterprise & support: Enterprise sales and custom implementations for large deployments (VPCs, compliance reviews, NDAs, etc.), active community/forum and documentation, and case studies/customer mentions (example: DigitalOcean referenced). Company background: public leadership/founders referenced (e.g., Daniel Vassilev, Jacky Koh), venture-backed (investors cited include Bessemer Venture Partners, Insight Partners, King River, Galileo Ventures, Archangel Ventures), and company size referenced on site (~80 in company copy). Notes & limitations: The pricing page and plan docs are the authoritative sources for exact limits and costs (the summary captures structure and key rules but not per-plan allowances or per-action pricing). Billing nuances (e.g., Vendor Credits roll over while Actions typically do not) and BYOK behavior may vary by plan and should be confirmed for specific customer scenarios. For enterprise deployments, request SOC 2 attestation, DPAs, and custom retention/data residency guarantees in writing.