AI Assistant skills are powerful tools within the suggested responses feature, designed to enhance your human agents' performance by providing real-time, contextual guidance. The term "skill" is widely used in Webex Contact Center to describe human agent proficiencies. To effectively leverage AI Assistant skills, it's crucial to understand their distinct nature and how to strategically deploy them to complement your human workforce.

This article provides strategic guidance for administrators on how to effectively deploy AI Assistant skills. It clarifies key terminology, offers best practices for linking skills to queues in various scenarios, and provides insights into designing skills for specific use cases.

In Webex Contact Center, the term "skill" can refer to different concepts. To effectively deploy suggested responses, it's crucial to understand what an AI Assistant skill is, and how it differs from other established terminologies:

• AI Assistant skill: A configurable entity within AI Studio that guides a human agent by providing real-time suggestions (information) and actions during customer interactions. It acts as an intelligent assistant to your agents.

  • Key purpose: To assist human agents in real-time by providing contextual suggestions and actions.

• Human agent skill (or agent skill): Refers to the proficiencies or attributes of a human agent (example, language fluency, product knowledge, technical expertise). These are used by the routing system to match customer interactions to the most qualified human agent.

  • Key purpose: To route customer interactions to the most appropriate human agent based on their capabilities.

• AI Agent: An autonomous entity also configured in AI Studio that directly interacts with customers (example, a chatbot or a virtual agent in an IVR). AI Agents handle interactions independently before potentially escalating to a human agent.

  • Key purpose: To autonomously handle customer interactions without human intervention.

Key distinction: An AI Assistant skill assists a human agent. A Human Agent skill describes a human agent's capability. An AI Agent interacts with a customer.

For more detailed definitions of these and other AI concepts, refer to the AI terminology and concepts in Webex Contact Center article.

AI Assistant skills are assigned to queues, ensuring that agents handling interactions from that queue receive relevant suggestions. While the procedural steps for linking are covered in Link AI Assistant skills to queues article, here are strategic best practices for deciding which skill to assign to which queue:

• One queue, one AI Assistant skill: A single queue can be linked to only one AI Assistant skill at a time. This ensures consistent and focused guidance for agents handling interactions from that queue.

• Queue specialization:

  • Highly specialized queues: For queues handling very specific topics (example, "Billing Inquiry Queue," "Technical Support - Product X"), assign a highly specialized AI Assistant skill designed precisely for that topic (example, "Billing Assistant Skill," "Product X Support Skill"). This provides deep, relevant guidance.

  • General purpose queues: For broader queues that handle a wider range of common inquiries (example, "General Customer Service"), assign a general-purpose AI Assistant skill that covers frequently asked questions and common agent tasks.

• Scenario: Queues with multiple human agent skills:

  • Even if a single queue is staffed by human agents with diverse human agent skills (example, "English Support," "Spanish Support," "Technical Support"), the AI Assistant skill assigned to that queue should be designed to support the overall purpose of that queue. The AI Assistant skill provides suggestions based on the conversation context and the queue's function, not necessarily mirroring the individual human agent's specific skill set.

  • Best practice: Design the AI Assistant skill to be broad enough to assist any human agent handling an interaction in that queue, regardless of their individual human skills. Language support for the AI Assistant skill should align with the languages handled by the queue.

Designing an effective AI Assistant skill involves more than just configuration; it requires strategic thinking about the specific operational scenario it will support. This section provides guidance on how to approach the design of your AI Assistant skills for common use cases.

For detailed steps on creating and managing AI Assistant skills, refer to the Create and manage AI Assistant skills article. For best practices on crafting instructions, structuring knowledge bases, and defining actions, refer to the Best practices for optimizing AI Assistant skills article.

• Product Support use case:

  • Goal: To assist agents in providing troubleshooting steps, product specifications, and warranty information for specific products.

  • KB content: Detailed product manuals, FAQs, common troubleshooting flows, warranty policies.

  • Actions: Suggesting "Retrieve Product Specs," "Initiate Troubleshooting Flow," "Check Warranty Status."

  • Instructions focus: Guide the AI to identify product names, symptoms, and provide step-by-step solutions or relevant documentation links to the agent.

• Sales Inquiry use case:

  • Goal: To help agents provide accurate product features, pricing, and qualify leads.

  • KB content: Product catalogs, pricing sheets, feature comparisons, lead qualification criteria.

  • Actions: Suggesting "Provide Pricing," "Check Stock Availability," "Create Lead."

  • Instructions focus: Guide the AI to identify customer needs, product interest, and suggest relevant sales collateral or next steps to the agent.

• Billing and Account Management use case:

  • Goal: To assist agents in handling payment processes, account updates, and common dispute resolutions.

  • KB content: Billing policies, payment methods, account update procedures, dispute resolution flows.

  • Actions: Suggesting "Process Payment," "Update Account Details," "Initiate Billing Dispute."

  • Instructions focus: Guide the AI to identify account numbers, transaction details, and suggest relevant procedures or actions to the agent.

• Scope definition: Clearly define what the AI Assistant skill will and will not cover. Avoid making the skill too broad, as this can dilute its effectiveness.

• Knowledge Base alignment: Ensure that the Knowledge Base linked to the skill is rich with accurate, relevant, and agent-centric information specific to the skill's defined scope.

• Action integration: Identify key actions that can genuinely automate or streamline agent tasks within that use case. Prioritize actions that reduce manual effort or ensure compliance.

• Instructions nuance: Tailor the instructions to the specific needs of agents handling that use case. Consider their typical workflow and the information they need most urgently.

By strategically deploying and designing your AI Assistant skills, you can maximize their impact on agent efficiency and customer satisfaction across your Webex Contact Center.