AI Agent Index
ByHeather MacAvelia·Independently reviewed·Published Mar 19, 2026·Updated Sep 13, 2026
Independently verified against live vendor data on Sep 2, 2026.

Enterprise AI agent platform for building and managing an AI workforce. 1,000+ native integrations, MCP server and client, SOC 2 Type II. Sales-led pricing, self-serve signup paused.

How we scored it

Autonomy

4/5

Integrations

5/5

Pricing clarity

2/5

Evidence

3/5

Setup

2/5

The facts

Relevance AI is an AI agent platform built for businesses to design, deploy, and manage an AI workforce of autonomous agents, on the metaphor that companies should hire, train, manage, and scale AI agents like employees. Founded in 2020 in Australia by Daniel Vassilev, Jacky Koh, and Daniel Palmer, and operated by OnSearch Pty Ltd, it has particular strength in go-to-market automation across sales, marketing, and customer success. The platform targets revenue teams, operations leaders, and founders who want to delegate whole business functions to AI rather than automate individual tasks. Relevance AI connects to CRM (Salesforce, HubSpot, Pipedrive, Zoho), sales engagement (Outreach, Apollo, Instantly, Salesloft), email and calendar (Gmail, Google Sheets, Google Drive), databases and warehouses (Airtable, Notion, Snowflake, Databricks), and communication tools (Slack, Microsoft Teams, WhatsApp Business). The vendor states 1,000+ native integrations on its product pages and in its llms.txt, 1,800+ on its integration pages, and 2,000+ on its pricing page. It runs on third-party models from OpenAI, Anthropic, Google Gemini, Mistral, and others, with bring-your-own-key and BYO LLM available. MCP runs in both directions. A documented MCP server and a Claude Code plugin let Claude Code, Claude Desktop, Cursor, VS Code, ChatGPT, and OpenAI Codex build and manage Relevance agents programmatically. An MCP client lets Relevance agents connect to external MCP servers, through preset connections for Notion, Canva, and Atlassian or any custom server URL. The visual builder is aimed at operators, with a code-first path through the MCP server and plugins for GTM engineers. The linked pricing page lists a single Enterprise plan sold through sales, with no published rates, and the signup page states that self-serve plans are being re-configured and that signing up on your own is temporarily disabled. The product documentation still describes a Free plan at USD 0, a Pro plan from USD 19 per month billed annually, a Team plan from USD 234 per month billed annually, and a custom Enterprise plan. Annual billing is stated as 33 percent below the monthly rate. The self-serve ladder is documented but was not purchasable at verification. Plans meter Actions and prepaid Vendor Credits for third-party model usage, with top-ups at USD 80 per 1,000 Actions and USD 20 per 10,000 Vendor Credits, and Vendor Credits roll over indefinitely while a subscription is active. Enterprise features include SSO and SAML, RBAC, audit logs, agent evaluations, A/B testing, calling and meeting agents, enterprise triggers for Salesforce, Snowflake, and Zendesk, and a dedicated account manager. What the entry point buys is a sales conversation. SSO is stated as an Enterprise feature and enterprise app triggers are Enterprise-only. The documented Free and Pro plans exclude end users, shared projects beyond one, calling and meeting agents, and agent evaluations. Relevance AI suits enterprise GTM and operations teams that want a managed multi-agent platform with governance built in. Teams that want to build and self-host their own automation should look at n8n. Teams that want a no-code assistant with published self-serve pricing should look at Lindy. Teams whose priority is breadth of app connections for two-app automations should look at Zapier or Make.com. Teams that want browser-native automation should look at Bardeen. Current state Q3 2026: Relevance AI offers Invent, a plain-language builder that generates agents, tools, and evals from a description, and evals that score agent quality continuously against expert-defined criteria. Three build paths are available: Invent, drag-and-drop, and MCP with a Claude Code plugin. Named customers include Canva, Autodesk, KPMG, Lightspeed Commerce, Rakuten Advertising, Databricks, Confluent, Freshworks, Aveva, Employment Hero, ThoughtSpot, Qualified, Send Payments, and Zembl. Published case studies state that Qualified generated USD 7M of pipeline with 35-plus agents, Send Payments saves 40 hours per week, and Zembl drives 30 percent more customer conversions. Enterprise security includes PII masking, SAML 2.0 SSO, audit logs, version control, and AWS data residency in the US, EU, or Australia. The trust center states that no customer data is used for training models. Relevance holds SOC 2 Type II with GDPR compliance, and lists an ISO 27001:2022 audit engagement with no certificate published. The company raised a USD 24 million Series B in May 2025 led by Bessemer Venture Partners, after a USD 10 million Series A in December 2023. It has also launched Arg, a separate free product for agents and teams working in a shared filesystem.

Pricing

custom

View pricing ↗

Segment

b2b

Setup

moderate

Verified

Sep 2, 2026

Transparency

Quote Only

Contract

Monthly or Annual

Data training

Not Trained

Human in loop

Optional

Capabilities

workflow-builderoutbound-automationlead-generationno-codecrm-syncautonomousscheduling

Pros & Limitations

Editorial assessment

Pros

  • ✓Invent builds agents from plain language descriptions: domain experts describe what they want and the platform creates agents, tools, and evals automatically, with three build paths (Invent, drag-and-drop, MCP) serving operators through to GTM engineers.
  • ✓Broad connectivity for GTM work: 1,000+ native integrations by the vendor's own count, an MCP server with a Claude Code plugin, and an MCP client that connects agents to external MCP servers such as Notion, Canva, and Atlassian.
  • ✓SOC 2 Type II with enterprise security depth: PII masking, SAML 2.0 SSO, audit logs, AWS data residency in the US, EU, or Australia, version control, and human-in-the-loop approval gates, with a stated no-training-on-customer-data policy.

Limitations

  • –The linked pricing page publishes no rates and every plan is sold through sales, and self-serve signup is stated as temporarily disabled while the vendor re-configures its self-serve plans, so costs cannot be evaluated without a sales conversation.
  • –A small G2 review base limits peer-validation depth: the published case studies (Qualified, Send Payments, Zembl) are strong, but procurement teams that weight review volume heavily will find less independent evidence than on larger platforms.
  • –AI workforce positioning requires meaningful investment to deliver value: realizing the productivity gains requires significant time spent designing, training, and evaluating agents, which operations teams may not absorb for casual use cases.

Technical Details

Deployment
webapi
Model architectureMulti-model (third-party LLMs, BYOK)
Autonomous rateConfigurable: agents run autonomously within configured quotas and guardrails; users approve workflow logic and review outputs
Integrations
HubSpotSalesforcePipedriveZoho CRMOutreachApolloInstantlySalesloftGmailGoogle SheetsGoogle DriveAirtableNotionSnowflakeDatabricksSlackMicrosoft TeamsWhatsApp BusinessLinkedIn AdsGongIntercomZendeskFreshdeskGitHubConfluenceCanvaWebflowTwilioPostHogActiveCampaignAhrefsBuiltWith
Security
SOC 2 Type IIGDPRCCPA

Similar agents

Rating

3.8/ 5
★★★★★

Editorial score

How we score this →

Recognition

MCP Server VerifiedListed 2026
Embed these on your site →

Industries

RetailMarketingConsultingEnergyFinance

Built for

B2BSaaS

Leave a review

Never displayed publicly.

Agent Stacks

See workflow stacks that feature Relevance AI.

Compare

Related Content

Is this your tool?

Claim this listing free to get your embeddable badges, a review collection link, and the Claimed badge on your listing.

Claim listing →

Free · Every Two Weeks

AI Agent Price & Rating Tracker

Price changes, new agent launches, acquisitions, and rating updates across the AI agent market. Verified against live vendor data, not vendor marketing.

No spam. Unsubscribe anytime. We never share your email.