Customers expect swift, accurate, and personalized service. To meet these demands, companies are increasingly turning to Generative AI to enhance their customer service capabilities. Unlike traditional AI, which is often rule-based and rigid, Generative AI can understand the context and generate a response on demand, such as text, art, and even code.

Generative AI has become indispensable. E-commerce platforms use it to automate customer support interactions, banks employ it for fraud detection and to provide personalized financial advice, and tech companies rely on AI for troubleshooting customer queries and extend self-service abilities. The common theme across these applications is the ability to empower support teams to deliver faster, accurate, and empathetic responses, making it an ideal solution for the modern customer service landscape.

Zia GenAI - Zoho's native Generative AI model

Zia, Zoho's built-in AI tool, wears multiple hats, from predictive abilities to automating crucial ticket handling actions to providing 24X7 self-service to customers. Zia can help support agents handle tickets with ease, provide them custom solutions, and resolve tickets faster without burnout.  

Zia GenAI is Zoho’s own Generative AI model that pairs analytical and generative abilities to enhance the self-service experience for customers and employees. The decision to build our generative model stems from the following observations:

• Privacy and security - Building our own model gives us an edge over maintaining data privacy and securing customers' data. Unlike OpenAI/ChatGPT, where the data goes out of Zoho's servers to generate responses, Zia GenAI ensures that all data stays within the Zoho ecosystem for Generative AI capabilities. This approach guarantees that customer data remains securely contained within Zoho's infrastructure, leveraging the Knowledge Base (KB) to analyze and generate suitable responses without compromising privacy or security.
  
  
• Secure data storage - Zoho's AI model does not store customer data in its servers. However, it processes the data temporarily to generate responses, ensuring that the data remains secure and is not retained beyond the response generation process. Zia GenAI never trains on customer data and exclusively only on the knowledge base articles to analyze and understand the context. Based on which, it reads the ticket and summarizes a response that best suits the query.
  
  
• Compliance adherence - Zia GenAI, like all Zoho services, strictly adheres to GDPR guidelines. The Zia GenAI model is hosted on Zoho's servers within the data center corresponding to the customer's region, ensuring that data never leaves the Zoho ecosystem. Additionally, the model is not trained on customer data, further safeguarding privacy and compliance.

Permission Required
Users with the ZIA permission under the Administrative section of their profile can configure Zia GenAI. 
Check Feature Availability and Limits

Prerequisite: Zia GenAI will only generate responses if the Answer Bot is configured for the department where you want the replies to be generated.

Points to remember

• Ensure that Answer Bot is configured for the appropriate department for the reply assistance feature to work in Zia GenAI.
• It is important to Review Zia GenAI’s suggestions before sending them to customers to maintain personalized support standards.
• For best results, regularly update the knowledge base to maximize the quality of AI-generated responses.

Enabling Zia GenAI is a simple and straightforward process. Follow the below steps to get started:

1. Authentication: Begin by enabling Zia GenAI within Zia powered by GPT. Unlike the ChatGPT option, there’s no need to provide an external API key, making the setup process hassle-free.
   
   
   
   
2. Feature preferences: Once enabled, the following features can be activated to enhance customer support operations:
   
   
1. Ticket Tone Analysis: Identifies the sentiment in a customer's message by analyzing its language and context. Zia GenAI assigns tone labels, such as happy, angry, sad, or frustrated, enabling agents to better understand the customer's emotional state.
   
2. Ticket Summary: Produces summaries for incoming inquiries and previous threads within a ticket. Options include recent, initial, or multiple ticket threads (up to 30, covering incoming, outgoing, forwarded, and public comments). This allows agents to quickly grasp the context and history of complex tickets, ensuring informed and relevant responses.
   
3. Reply Assistance: Provides suggested responses for the latest incoming customer inquiries using only the knowledge base articles. Responses can be either auto-generated or generated when agents click the Predict button, ensuring that replies are accurate and consistent with company policies.
   
   
   
   
3. General preferences: In General Preferences, select the desired response preferences to choose between auto-generated or manually generated responses. When auto-generation is enabled, Zia GenAI automatically provides ticket summaries, tone analysis, and response suggestions. If not, agents can manually generate them by clicking the Generate button for each ticket.
   
   
   
   

Note:

• The availability of the Ticket Tone Analysis, Ticket Summary, and Reply Assistance options in the ticket detail view will depend on the features that were selected during the configuration Zia GenAI's feature preferences.
• Users can disable Zia auto-suggestions in the iBar when they click "Reply All" by navigating to Setup > Personalization > Preferences and switching off Auto Suggestions. This will stop Zia from opening automatically.
  

Prerequisite: To make use of the Zia GenAI, an Answer Bot must be created for the department to which the ticket belongs.

1. Navigate to Setup > Zia > Answer Bot.
2. In the Answer Bot page, choose a Department and click Add Answer Bot.
3. In the Create Answer Bot page, provide the following details in Answer Bot details.
   
   1. Enter a Bot name.
   2. Add a description.
   3. Enable the Bot either for agents or customers or both.
   4. If enabled for customers, select the channels where the Bot will be deployed.
   5. Enable Generative AI integration.
   6. Set up the Default Messages.
4. Click Save.

For further details on Answer Bot, please refer to the help article on AI-based Self-Service with Answer Bot.

Accessing Zia GenAI from Tickets 

Once Zia GenAI is enabled, agents can access it within the ticket detail view. In the ticket detail view, Zia GenAI enables agents to perform the following actions:

 Ticket Intelligence

 Zia GenAI summarizes incoming customer inquiries and detects the emotions conveyed in the message. This saves agents time and effort by providing a concise overview of the customer's concern, enabling them to address issues more effectively. Additionally, it identifies the customer’s tone and sentiment, such as frustration or happiness, helping agents provide empathetic and personalized responses.

For example, consider a customer contacting a software company about a recurring issue with their product. Zia GenAI’s Ticket Prediction summarizes the customer’s inquiry, highlighting key details of the problem and their frustration. This equips the agent with a clear understanding of the issue, allowing them to respond efficiently and empathetically, while reassuring the customer that the technical team will investigate thoroughly.

Reply Assistance  

The Reply Assistance feature helps agents generate optimal replies for customer inquiries using the organization’s knowledge base. This ensures responses are accurate, consistent, and aligned with company policies. Here’s how it works based on the knowledge base configuration:

• Relevant articles found:
  
  • Zia GenAI generates a response using the relevant knowledge base article.
  • The article source is specified, and agents can copy the content into the reply editor or add it as a comment within the ticket.
    
• No relevant articles found:
  
  • Zia GenAI notifies the agent that no related articles are available and prompts the team to add relevant knowledge base content for future use.
    
    

For example, in a tourism-related inquiry where a customer asks for trip suggestions, Zia GenAI generates a response using knowledge base articles about local attractions, outdoor adventures, or historical sites. If the knowledge base lacks relevant information, Zia GenAI suggests adding the necessary content, ensuring future inquiries can be addressed more effectively.

Ask a Question  

With the Ask a Question feature, agents can query Zia GenAI directly to obtain responses based on the configured knowledge base. This is useful for scenarios requiring quick access to specific information.

For example, if an agent asks for recommendations on tourist attractions in a city, Zia GenAI provides a response based on relevant knowledge base articles. If no articles exist, it advises updating the knowledge base to include this information.

If the ticket prediction or response generated by Zia GenAI does not meet expectations, it can be regenerated until a satisfactory result is achieved. In the Reply Assistance tab, agents have the option to select from various tone styles, including Formal, Informal, Diplomatic, Assertive, or Humorous. Below is a brief description of each tone:

• Formal: A professional and polite tone, ideal for business communications and scenarios requiring a respectful approach.
• Informal: A casual and friendly tone, suitable for relaxed interactions and conversational exchanges.
• Diplomatic: A tactful and considerate tone, designed for handling sensitive matters while maintaining positive relationships.
• Assertive: A confident and direct tone, appropriate for clearly conveying positions or requirements respectfully.
• Humorous: A light-hearted and engaging tone, aimed at creating enjoyable and relatable interactions.
  
  

Additionally, agents can adjust the response length as needed, selecting either a longer or shorter response to best suit the context of the conversation.

To access Zia GenAI from tickets

1. Navigate to the Tickets module.
2. Open the desired ticket in its detail view.
3. Click the Zia Powered by GPT icon (  ) in the left pane.
4. In the Zia Powered by GPT window, do the following:
   1. Under Generative AI section, Open the Ticket Tone and Ticket Summary tab to get the ticket summary and ticket tone analysis of the latest incoming customer inquiry.
   2. Open the Reply Assistance tab to get an automated optimum response. Click Use as Reply to use the generated response as the ticket reply or comment.
5. Click Send when you're satisfied with the ticket prediction and response.