Docs

Documentation

Automation Integrations

No-code AI automation. Trigger ModelRiver requests from form submissions, Slack messages, CRM updates, and 6,000+ other apps: without writing a single line of code.

Overview

Not every AI-powered workflow needs custom code. Automation platforms like Zapier connect ModelRiver to thousands of apps, allowing you to build AI-powered automations through a visual interface.

Why use ModelRiver with automation tools?

  • No code required: Build AI workflows visually
  • 6,000+ integrations: Connect to virtually any SaaS tool
  • Same reliability: Automatic failover and cost tracking for every trigger
  • Structured outputs: Get machine-readable JSON responses for downstream automations

Supported tools

ToolTypeHighlightsDifficultyGuide
ZapierNo-code automation6,000+ apps, webhooks, multi-step Zaps⭐ EasyView guide →

Zapier

Build multi-step AI automations by connecting ModelRiver to any of Zapier's 6,000+ apps. Common use cases:

  • Summarise form submissions: Google Forms → ModelRiver → Gmail
  • Auto-reply to Slack messages: Slack → ModelRiver → Slack
  • Enrich CRM contacts: HubSpot → ModelRiver → HubSpot
  • Translate support tickets: Zendesk → ModelRiver → Zendesk
JSON
1{
2 "workflow": "my-chat-workflow",
3 "messages": [
4 { "role": "user", "content": "{{trigger_data}}" }
5 ]
6}

Full Zapier guide →

Tips for automation workflows

  1. Use structured outputs: Configure your ModelRiver workflow to return JSON, making it easy to extract fields in downstream steps
  2. Use wrapped format: Set "format": "wrapped" to get metadata alongside the response
  3. Keep prompts short: Automation triggers should use concise, focused prompts to minimise token costs
  4. Monitor usage: Check Request Logs to ensure automation-triggered requests aren't eating your budget

Next steps