<span id="top"></span>

Understanding How Piper the AI SDR Agent Handles PII

This article explains how Piper the AI SDR Agent interacts with visitor and lead information, including Personally Identifiable Information (PII), during conversations or interactions. Because you can customize your AI Agent's name and personality, we will refer to it as the 'AI Agent' throughout this guide for clarity.

• PII Handling in AI Agent Website Conversations
• PII Handling in AI Agent Email Conversations
• Visitor Control and Our Commitment to Data Security

<span id="website-conversations"></span>

PII Handling in AI Agent Website Conversations

Understanding how the AI Agent manages PII during live website conversations is essential for our Qualified users and stakeholders. The AI Agent is designed with a privacy-first approach to ensure data protection and customer trust. 

Here’s a breakdown of the AI Agent's default access, how it learns names, and how you can configure its behavior:

Default Information Access & How the AI Agent Learns Names

By default, for general website conversations, the AI Agent does not access a visitor’s first name or company information from your organization’s connected backend systems. Initially, the AI Agent will use generic terms to address the individual they are communicating with, however, the AI Agent can learn and use a visitor's name during a conversation in a couple of ways:

• Directly from the Visitor: If a visitor shares their name (e.g., "Hi, I'm Sarah"), the AI Agent will remember it from the current conversation history to maintain a natural dialogue.
• Continuity with Returning Visitors: For returning visitors who have been cookied, the AI Agent may refer to a name shared by the visitor in a recent previous conversation (from approximately the last 30 days of conversation history) to create a continuous experience.
• Last Names: The AI Agent only becomes aware of a visitor’s last name if the visitor provides it directly during a conversation.

Admin Configuration for "Pounces"

Qualified Admins have control over how the AI Agent uses visitor information in "Pounces" – those proactive messages the AI Agent sends to engage visitors when they arrive on the website. 

• Personalizing Pounces:  Pounces can be configured to use a visitor's first name and company details for highly personalized outreach.
• Default Setting for New AI Profiles: For all newly created AI Profiles, the ability for the AI Agent to use first name or company information in Pounces is turned OFF by default.
• Action for Admins: Review Existing AI Profiles: If your organization has existing AI Profiles, we strongly encourage  reviewing your Pounce configurations to ensure the AI Agent's behavior aligns with your company’s PII policies. You can find these settings within the Qualified app by going to Settings → AI Studio → Pounce.

For more information about Pounce settings, check out: Automating Visitor Engagement with Pounce

PII in Qualification Prompts

The information used in AI Agent conversations is sourced directly from the conversation the AI Agent is having with the website visitor. Even if the AI Agent asks for and captures information like a name or company name to populate a field during its qualification process in a conversation, this information is sourced from the live conversation history. The AI Agent does not pull PII from connected Salesforce or other data systems to inject into its general conversational prompts, outside of the specific Pounce configurations set by your organization.

Back to Top

<span id="email-conversations"></span>

PII Handling in AI Agent Email Conversations

Beyond real-time conversations, the AI Agent also extends its powerful capabilities to automated follow-up email campaigns.

Building Email Campaigns with the AI Agent

• Default Lead Information for Drafting: When building a new email campaign, the AI Agent is supplied with the lead's Full Name and Company fields by default from your organization’s lead data to help draft relevant follow-up emails.
• Full Admin Control and Customization: Qualified Admins have complete control. These default fields can be removed. You can also add to or customize the information used by the AI Agent by selecting other fields from your organization's lead records. This customization is key for effective and compliant email generation.

Handling Incoming Email Replies (the AI Agent Replies)

• Access to Reply Content: When a lead replies to an email from the AI Agent, the AI Agent accesses the content of that incoming reply to understand context and generate an appropriate automated response.

Handling Incoming Email Replies is currently in beta testing with a select group of customers. Please reach out to your designated Qualified Success Architect (QSA) to learn more.

Back to Top

<span id="our-commitment"></span>

Visitor Control and Our Commitment to Data Security

• Visitor Control: Ultimately, visitors always have control over the information they choose to share during a conversation with the AI Agent or in their email replies.
• Our Commitment to Continuous Improvement: At Qualified, we're always working to make the AI Agent smarter and safer, continuously enhancing its capabilities and commitment to data security to serve you and your customers better.

Back to Top