What is Relay.app?
Relay.app is a no-code workflow automation platform built for teams that need to automate without fully removing humans from the picture. Founded in 2021 and backed by Khosla Ventures and Andreessen Horowitz, Relay sits in the same category as Zapier and Make.com but takes a distinct position: it treats human approval steps as a first-class feature rather than an afterthought. Teams connect tools like Gmail, Slack, HubSpot, and Notion, then build workflows that can pause mid-run to wait for a human decision before continuing. By late 2025, Relay reached $3M in annual revenue — strong traction for a 27-person company, suggesting its design philosophy resonates with a loyal, self-serve user base.
Key Takeaways
- Human-in-the-Loop steps let workflows pause for approvals or decisions — no other major no-code tool makes this a first-class primitive.
- All plans include AI steps powered by GPT-4o without requiring your own OpenAI API key.
- The integration catalog covers 100+ tools — useful but meaningfully smaller than Zapier's 8,000+.
- Step limits reset monthly; at 200 steps on the free plan, most real workflows will outgrow it quickly.
- Relay reached $3M ARR with 27 employees, signaling product-led growth and a sticky, self-serve user base.
What Makes Relay.app Different
Relay's core insight is that most business processes aren't fully automatable — somewhere in the chain, a human needs to review an output, approve a decision, or add context that software can't infer. Where Zapier and Make treat automation as a straight pipeline from trigger to action, Relay builds pause-and-resume into the workflow model itself. A contract approval flow, for example, can auto-populate the document, route it to the right person for review, wait for their sign-off, then continue with the downstream steps. The AI integration reinforces this: instead of replacing human judgment, Relay's AI steps generate drafts or summaries that a human can then approve or edit before the workflow proceeds. It's automation with a co-pilot seat built in.
Relay.app vs Zapier vs Make
Think of these three tools as occupying different points on a spectrum from easy-but-limited to powerful-but-complex. Zapier is the safe enterprise choice — 8,000+ integrations, organizational governance, and a brand that IT teams recognize. Pick Zapier when coverage matters more than cost or when you need to justify the tool to a procurement team. Make.com earns its place at high volume: its per-operation pricing and visual flow-chart canvas reward teams with complex, branching data transformations who have hit Zapier's price ceiling. Pick Make when you're running thousands of operations monthly on a budget. Relay.app wins when the workflow itself involves human touchpoints — compliance reviews, client-facing approvals, content sign-off queues. One practitioner reported moving 80% of their Zapier automations to Relay, keeping the remaining 20% only because of specific missing integrations. That ratio captures Relay's position well: it wins on experience and HITL design, still loses occasionally on raw coverage.
Pricing
Relay.app's Free plan gives 200 automated steps and 500 AI credits per month — enough to evaluate the tool, not enough for an active workflow. The Professional plan at $38/month (or $19/month billed annually) bumps to 750 steps and 5,000 AI credits. The Team plan at $138/month adds shared app connections, up to 10 users, and 2,000 steps. Enterprise pricing is custom and requires contacting sales. A few things worth knowing before you commit: standard steps and AI credits are metered separately, so a workflow heavy on AI summarization will exhaust credits faster than a basic trigger-action chain. When steps run out, workflows pause until the monthly cycle resets — which can disrupt operations during a busy period. All plans get access to the same integration catalog, so the upgrade decision is purely about volume and collaboration features.
Who Uses Relay.app
Relay's sweet spot is operations-focused roles at small to mid-sized companies — RevOps, marketing ops, and customer success teams at 10–200 person organizations. The HITL design makes it a natural fit for compliance-sensitive processes (legal review gates, client approval steps) and client-facing workflows where a fully autonomous automation would feel too risky. On the freelance and fractional side, automation consultants increasingly recommend Relay to clients who have outgrown Zapier's simplicity but aren't ready for Make's steeper learning curve. It pairs naturally with the modern SMB stack: Google Workspace, Slack, Notion, HubSpot, and Airtable. At Pangea, we see Relay.app appearing more frequently in fractional RevOps and operations consultant role postings alongside Zapier and Make — rarely as a standalone requirement, but as a signal of fluency with AI-enabled automation tooling.
Real Limitations to Know
The integration gap is the most common reason teams keep a parallel Zapier account: 100+ integrations covers the mainstream SaaS stack well but leaves a long tail of niche apps unaddressed. There is no version control for workflows — if an automation breaks after an edit, rolling back requires manual reconstruction. As a seed-stage company with roughly 27 employees, Relay carries platform risk that Zapier or Microsoft Power Automate do not; it's worth factoring into decisions for business-critical automations. Finally, the dual metering of steps and AI credits adds billing complexity: teams need to monitor both limits independently, and AI-heavy workflows can exhaust credits well before the step limit is reached.
The Bottom Line
Relay.app has carved out a defensible niche in workflow automation by treating human judgment as a design requirement rather than a gap to eliminate. For teams running approval-heavy or compliance-sensitive processes, its Human-in-the-Loop architecture solves a real problem that Zapier and Make only partially address. The tradeoff is integration breadth and platform maturity. For companies hiring through Pangea, Relay.app expertise signals an automation specialist who understands that the best workflows aren't always the most autonomous ones.
