Skip to main content

AI Builder Improves Wait Step Generation for Complex Prompts

· 2 min read
SMBcrm Team
SMBcrm Team
Product Team

AI Builder: Smarter Wait Step Generation​

AI Builder now creates Wait steps more reliably and predictably, especially when your workflow needs multiple wait types and conditions built from a single prompt. This update also adds guardrails that reduce errors and keep your workflow structure intact.

What’s New​

Handle complex wait requests in one prompt​

You can now describe layered wait logic in a single request and have AI Builder generate the right sequence of Wait steps automatically.

Example prompt: “Add a wait with a delay of 3 minutes, then another with a delay of 10 hours, then another but only run it on Tuesday–Thursday between 10am–5pm, then add an SMS and then wait for reply with a timeout of 10 minutes.”

AI Builder can correctly generate and configure:

  • Time delay waits (minutes, hours, etc.)
  • Business hour window waits (days and time ranges)
  • Reply-based waits with explicit timeouts

What’s Fixed​

Safer wait times​

AI Builder now blocks negative or invalid values, normalizes time units, caps extreme values, and intelligently interprets overly large ranges (for example, “years”) to prevent unusable waits.

Predictable Wait type selection​

Wait step selection is now more consistent:

  • Simple delays and time window phrasing produce time-based waits
  • Timeout branches only appear when you explicitly ask for them

Protected workflow structure​

Time-based waits are enforced as a single path. If a wait is accidentally created as a multi-path node, AI Builder automatically corrects it so your workflow doesn’t end up with broken transitions.

Timezone and multi-path improvements​

  • Timezone fields populate correctly
  • Multi-path wait types consistently include configured timeout values

Why It Matters​

If you build automations regularly, Wait steps are often where small configuration errors can cause big timing problems. These improvements help you generate more accurate workflows from plain English, reduce manual cleanup, and keep automations running as expected, especially when you’re building multi-step sequences.