How Qvasa's AI Works: Qualitative Questions and Intelligent Insights – Qvasa

Qvasa’s AI capabilities are designed to deliver deep, actionable insights into customer support operations. Central to our AI is the concept of qualitative questions—specific questions that the AI answers for every ticket to extract meaningful information. These insights empower organizations to improve customer experiences, optimize operations, and better understand customer needs.

The Foundation: Categorization

Before activating AI features, users define 6–10 key categories that encompass their tickets, tailored to their business context. This initial setup ensures that AI outputs are relevant and actionable.

User-Defined Categories: These categories form the backbone of Qvasa’s AI. They are customized based on operational priorities and customer support needs.
Example: Categories such as "Billing Issues," "Technical Support," or "Product Feedback."
Contextual Training: Users provide examples and business-specific rules to help the AI accurately classify tickets.
Triggering Events: Categorization can run on specific Zendesk ticket status changes, such as when tickets move to "Closed" or "Solved," or custom combinations like "New and Closed.".

This structured setup ensures that subsequent AI workflows, such as Contact Reasons and Inferred CSAT, have a solid foundation.

Qualitative Questions: Fixed and Generative

At the heart of Qvasa’s AI are qualitative questions—specific questions that extract insights from tickets. These can be:

Fixed Questions: Users provide predefined options for the AI to choose from.
Example: "What category does this ticket belong to?"
Generative Questions: The AI generates its own responses based on ticket content.
Example: "Why was the customer upset in this ticket?"

These qualitative questions unlock powerful insights and drive several key features.

Key Features of Qualitative Questions

1. Nested Qualitative Questions

Qvasa supports nested questions, enabling a deeper level of insight.
How it works: If a ticket is tagged with a fixed category like "Login Issues," the AI can ask a follow-up question to further refine the categorization, such as tagging the ticket with "Sign-Up Issues" or "Log-In Issues."
This dynamic capability allows organizations to uncover granular details within broader categories.

2. Filtering by Ticket Fields

Qualitative questions can be filtered to run only on specific tickets.
Example: A question like "Why was the customer dissatisfied?" can be set to apply only to tickets that are closed, high-priority, or tagged with specific static or dynamic attributes.
This ensures that AI workflows remain focused, relevant, and efficient.

Core AI Workflows

1. Contact Reasons

When categorization is enabled, generative "Contact Reasons" are automatically activated.
For each ticket, the AI evaluates whether existing contact reasons apply. If none match, it generates a new contact reason.
Reportable in Dashboards: Contact reasons are stored as reusable, reportable fields. Users can track trends, frequencies, and patterns in Dynamic Dashboards.

Example Insight: Identify the top three reasons customers contact support over the past month.

2. Inferred CSAT

This is an out-of-the-box AI question applied to every closed ticket.
Question: "If the customer were asked to rate their satisfaction with the support they received, how would they respond?"
The AI categorizes the response into one of four buckets:
- Satisfied
- Neutral
- Dissatisfied
- Unknown

Value Add: Even without formal CSAT surveys, users can measure customer sentiment and identify areas for improvement.

3. DSAT Reasons

This workflow identifies the underlying reasons for dissatisfaction (DSAT). It is triggered when:
- A negative CSAT score is received.
- The Inferred CSAT workflow determines a ticket would have been rated as "Dissatisfied."
Reportable Field: DSAT reasons allow teams to pinpoint common causes of dissatisfaction and prioritize improvements.

4. Product Reasons

This feature categorizes issues into three distinct reportable fields, providing a detailed view of why tickets are created:

Product Deficiency Reasons:
When: The issue arises from an actual bug or quality problem in the product.
Example: A confirmed bug causing unexpected behavior.
Product Confusion Reasons:
When: The issue stems from customer misunderstanding of how the product works. These tickets are resolved with explanations or documentation.
Example: A customer thought a feature worked differently than it actually does.
Agent Intervention Reasons:
When: The issue required an agent to execute a workflow or action that the customer could not do independently.
Example: A password reset requiring back-end intervention.

By categorizing tickets in this way, organizations can target specific areas—whether improving product quality, clarifying documentation, or empowering customers with self-service tools.

How It All Works Together

Qvasa’s AI features are designed to work seamlessly with Zendesk data to deliver unparalleled insights:

Categorization is the Key:
All workflows, from contact reasons to product reasons, build on the foundation of user-defined categorization.
Reportable and Actionable Data:
The AI-generated insights are not just answers—they are fields that can be analyzed in Qvasa’s Dynamic Dashboards. Users can track trends, performance metrics, and root causes with ease.
Automated Yet Customizable:
While many workflows (e.g., Contact Reasons and Inferred CSAT) are automated, users retain control through customization of categories and qualitative questions.

Why Qvasa’s AI Is Unique

Customizable and Flexible: Users can tailor categorization and questions to fit their unique needs.
Reportable Insights: Every AI-generated field is designed for analysis and action.
Dynamic Workflows: Features like nested questions and filtering by ticket fields ensure precision and relevance.
Business-Ready Features: Out-of-the-box workflows like Inferred CSAT and Product Reasons deliver immediate value without requiring additional setup.

Qvasa’s AI transforms raw ticket data into structured, actionable insights. By focusing on qualitative questions and leveraging the power of categorization, contact reasons, inferred CSAT, DSAT reasons, and product reasons, our platform provides organizations with the tools they need to optimize customer support and improve operational outcomes.

Ready to unlock the potential of AI in your customer support workflows? Let Qvasa guide you to better insights today.

Related to