FWIW, the RONA timer doesn’t start when the call arrives at the agent - it starts when the agent’s browser receives the notification.
YMMV, but remote agents on slow connections see longer effective RONA times because the notification delivery is delayed.
We decided to implement PCI-DSS compliant payment collection inside Architect using Secure Pause.
The flow pauses the recording, the agent asks for the card number, the customer enters it via DTMF, and the flow resumes recording. Does the DTMF data get logged anywhere in the platform, or is it truly ephemeral?
From a compliance perspective, you must also disable screen recording during the Secure Pause.
If the agent’s screen shows the DTMF digits being entered (e.g., in an Architect script panel), the screen recording extension will capture them even though the audio recording is paused. This creates a PCI scope leak. Ensure your Chrome recording extension is configured to pause simultaneously.
We built a custom React payment form that overlays the agent’s screen during Secure Pause.
// Trigger overlay when Secure Pause starts
framework.on('conversationEvent', (event) => {
if (event.body.securePause === true) {
setShowPaymentOverlay(true);
// Overlay blocks screen recording from capturing DTMF
}
});
The overlay displays a generic ‘Payment in Progress’ message, preventing the screen recording from capturing any digits.
In PureConnect, we handled this using Interaction Recorder’s built-in PCI pause functionality.
The IC Recorder had a native API call that paused both audio AND screen recording simultaneously with a single command. GC’s architecture requires you to coordinate the audio pause (via Architect) and the screen pause (via the browser extension) separately. It is more fragile.
DTMF digits entered during Secure Pause are passed to the Data Action target but are NOT stored in the conversation record.
The Architect flow captures the DTMF input, passes it to your PCI-compliant payment gateway via a Data Action, and the digits exist only in transit. They are not written to the analytics database, the recording metadata, or the flow execution logs - provided you have the ‘Secure’ flag enabled on the DTMF variable.
We are configuring email routing, but HTML email templates render differently across email clients.
Our canned response templates look perfect in Chrome but are completely broken in Outlook 2019. The CSS flexbox and grid properties are stripped by Outlook’s Word-based rendering engine. We had to rewrite every template using inline styles and <table> layouts.
We migrated from Zendesk’s email template system to GC, and the template limitations were the biggest surprise.
Zendesk has a WYSIWYG template editor with variables. GC’s canned response system is much more basic. You can use substitution variables like {{participant.name}}, but complex conditional logic (e.g., ‘show this paragraph only for VIP customers’) requires a Data Action to pre-render the email body externally.
# Terraform - Deploy canned responses as code
resource "genesyscloud_routing_email_domain" "support" {
domain_id = "support.example.com"
}
resource "genesyscloud_quality_forms_evaluation" "email_qa" {
name = "Email Quality Form"
published = true
}
Manage email routing domains and QA forms via Terraform. Don’t click the UI.
Does email routing support automatic language detection?
We receive emails in English, Spanish, and Portuguese. Can the platform detect the language of the inbound email and route it to the appropriate language-skilled queue automatically, or do we have to build a Data Action that calls a translation API first?
For our Japanese enterprise deployment, email routing has a specific character encoding requirement.
Inbound emails from Japanese carriers (docomo, au, SoftBank) sometimes use ISO-2022-JP encoding instead of UTF-8. The GC email parser occasionally corrupts the subject line, replacing kanji characters with question marks. We filed a support case and Genesys confirmed it is a known issue with a planned fix.