Make: Visual Automation and AI Agents

Make (formerly Integromat) is a visual, no-code automation platform that connects more than 3,000 pre-built apps on a canvas of scenarios. You drag modules onto a canvas, connect them with lines, and add routers, filters, and loops for branching logic that a simple trigger-action tool cannot express. Each action in a scenario consumes a credit (usually one per operation), and plans scale by monthly credit allowance, which is why the platform is often described as the value play against Zapier: more complex workflows per dollar, with a steeper learning curve.

A scenario typically triggers on data from one app (for example, a new row in Google Sheets), transforms it, and sends it to another app. Make Grid gives a live map of every scenario, agent, and app so teams can see the whole automation landscape at once, and the analytics dashboard tracks utilization over time.

The platform has added AI-native pieces around that same canvas. Maia by Make builds automations and agents from a natural-language prompt, Make AI Agents add reasoning and tool use across the connected apps with a transparent reasoning panel on the canvas, and a cloud-hosted MCP server lets any AI interface invoke scenarios. More than 400 pre-built AI apps are available, and you can run custom JavaScript or Python with the Make Code app.

For our lab, the relevance is as a no-code counterpart to code-centric harnesses: where Claude Code or Kepler run agents in worktrees, Make runs them on a visual canvas with shared team visibility. It is enterprise-ready (GDPR, SOC 2, SSO, role-based access) and part of Celonis since 2020, so the trade is control and visibility for a learning investment before the first scenario runs clean.