Automation & Agents
Workflow automation and autonomous agent platforms.
Featured
No tools match these filters right now.
All Automation & Agents tools

MailerooMaileroo Group Pty Ltd
The Platform Built for Better Email Delivery
From $10/mo

Customizable agentic assistants for your work, with you approving every output
Freemium

One web context API for software and AI agents
From $25/mo

No-code automation across thousands of apps
From $19.99/mo

Visual, canvas-based automation for complex workflows
From $9/mo

Open-source workflow automation with a code escape hatch
Freemium

Custom AI chatbot builder trained on your data
From $40/mo

AI accounting automation and receipt extraction
From $32/mo

No-code AI workflow and web scraping platform
From $37/mo

Enterprise sales execution and revenue intelligence
Contact for pricing

Generative AI sales sequencing and prospect research
From $49/mo

AI productivity workspace and autonomous agents
From $10/mo

AI financial close and accounting automation for startups
Contact for pricing

Collaborative conversational AI agent builder
From $60/mo
Questions about automation & agents tools
Zapier, Make or n8n?
Zapier for the widest integration catalogue and the shortest path to something working. Make when the workflow branches enough that a diagram beats a list of steps. n8n when you want to self-host, write a real code step, or stop paying per task at volume — it is source-available, which is a different risk profile from a closed platform holding your workflows.
What happens when a workflow fails at three in the morning?
This is where the platforms actually differ. Make treats error handling as part of a scenario and n8n has dedicated error workflows, so a failure gets caught and retried. A chain built without either does the first three steps, fails the fourth, and tells nobody.
Do I need an automation platform to use AI in a workflow?
No, but it is usually the cheapest way to try. All three have AI steps that summarise, classify or extract inside a larger flow, which lets you put a model in the middle of a real process without standing up a service to hold it.