Skip to main content
Signal AI recognizes patterns in your calls to report business outcomes in real time, trained on your own call recordings.
What are the most important insights your business can learn from your calls, and how can you capture them in real time? Signal AI automatically recognizes patterns in your calls to report success events — such as when a purchase is made or a quote is requested. Because Signal AI is trained by you on your own calls, it can accurately predict the kinds of call insights your business cares about most. A few reasons you might choose to capture real-time call insights with Signal AI:
  • You need your call insights more quickly than other systems can provide. Sending call results manually from your call center to your marketing software can take weeks — by then, any insights are too slow to inform quickly-changing marketing efforts.
  • You want to automatically adjust your other marketing technologies. Connecting Signal AI’s real-time insights into your other marketing software can automate keyword bidding, content personalization, and audience seeding and suppression.
  • You want to understand your customers better. Signal AI insights customized to your business and based on your own calls help marketers identify consumer intent quickly and clearly.
Before you begin: you’ll need audio recordings of a large sample of your calls. Your Invoca account may already record calls driven to you using Invoca promo numbers — if available, these are the best recordings to use.
  • If you have call recordings in your Invoca account, skip to “Creating custom Signal AI rules for your business” below.
  • If your recordings live elsewhere, see Signal AI Training: Recording Imports.
  • If you don’t have recordings at all, consider applying Signal AI insights based on your industry for now — see Industry Signals: Pre-Configured Insights — and return to this article once you’ve accumulated sample data.
If you also need to capture Signal AI insights in languages other than English, see Capture Signal Insights: Multi-Language Support.

​
Creating custom Signal AI rules for your business

Once you’ve identified your sample of call recordings, define what business insights you want Signal AI to look for.
  1. Log in to your Invoca account. In the sidebar menu, click Signal, then select Manage Signals.
  2. Click + New Signal, then select AI Signal, then Custom AI Signal.
  3. Enter a name for your Signal based on the outcome you want it to track. Add a more detailed description if needed.
  4. If the Signal represents a revenue opportunity, enter the expected revenue amount in the Revenue field.
  5. Choose whether this Signal applies to every campaign or only specific campaigns, using the “Apply This Signal” radio buttons. If applying to specific campaigns, choose them in the menu that appears.
  6. Select the Future Campaigns checkbox if you’d like this Signal to apply to every subsequent campaign you create. Otherwise, leave it unchecked.
  7. If you use syndicated campaigns and want to share this Signal with syndication partners, select the Syndication checkbox.
  8. Click Save.

​
Training your Signal AI

Now that you’ve identified a sample of recordings and created a Signal, the next step is training Signal AI to identify patterns and recognize success events in future calls. Report the results of your Signal for each call in your sample — called ground truth data — so Signal AI can identify patterns in successful and unsuccessful calls, then apply those patterns going forward. Follow the guide on Improving Signal AI Accuracy to provide ground truth data. Once training is complete, you’ll have a customized Signal that feeds real-time call insights into your marketing efforts. To continue building out real-time call intelligence, repeat these steps to create more customized Signals, connect your Invoca account to another system to pass insights along in real time, or keep refining accuracy by scoring more call recordings.

​
Where to go next

Last modified on September 23, 2026