# Custom Prompting **Fully Tailor Your Voice AI’s Interaction Style, Logic, and Response** At the core of NexaVoxa's conversational intelligence is the **customizable prompting** capability, which gives businesses the ability to shape every aspect of how their voice agents speak, react, and process information. Prompts define the **behavior** and **tone** of AI voice interactions, and NexaVoxa allows you to go beyond simple dialogue scripting enabling intricate, real-time adaptability for your agents. Custom prompts are an essential tool in creating human-like, brand-consistent, and business-specific voice agents that meet your unique objectives, whether for **sales calls**, **customer support**, or **automated appointments**. In NexaVoxa, **prompting** refers to the direct instructions given to the AI agent to guide its conversation flow. These instructions inform the AI on how to respond to specific user inputs, what tone to adopt, and how to handle particular situations during the call. Prompts are flexible and can be fine-tuned for: * **Tone of voice**: Make the agent sound friendly, formal, empathetic, etc. * **Intent-based responses**: Direct the agent to respond based on the user’s question or intent. * **Context handling**: Adjust agent behavior according to specific data about the user (e.g., their previous interaction history or preferences). #### Key Customization Features **Tone Customization** The tone of the conversation can be **adapted to fit your brand’s voice**, ensuring that the AI speaks in a manner that aligns with your business goals and audience expectations. Whether your brand demands a **professional**, **friendly**, **casual**, or **empathetic** tone, NexaVoxa’s prompting system provides the flexibility to modify how the agent communicates across every touchpoint. * **Example**: In a healthcare scenario, your AI agent might use a warm, compassionate tone when scheduling a doctor's appointment, while it may adopt a more formal tone when handling insurance claims for a financial services firm. **Domain-Specific Language** NexaVoxa allows you to **embed industry-specific terms** into prompts, ensuring that the agent uses the correct terminology based on your sector. This is particularly useful for industries such as **finance**, **legal**, **healthcare**, and **real estate**, where accurate jargon and phrases are essential for clear communication. * **Example**: In a financial services flow, the AI agent might reference terms like **“credit score,” “loan application,”** or **“interest rates”**, while in a retail environment, the agent might discuss **“order numbers,” “shipment tracking,”** or **“product categories.”** **Decision Logic and Behavior** Prompts can include **logical rules** to define how the agent behaves at different stages of a call. These can range from simple instructions like “repeat the question if the answer is unclear” to complex decision-making logic based on user input, data in your CRM, or past interactions. * **Example**: If a customer asks about a specific product and the agent detects uncertainty, a decision rule could trigger the agent to **escalate** the query to a human agent, ensuring that the customer’s issue is addressed correctly. **Rule-Based Logic (Constraints and Fallback Instructions)** You can define rules within the prompt to set **constraints** around how the AI agent responds. For example, an agent can be **restricted from making specific recommendations** unless a user has been **verified** or **qualified**. * **Fallback instructions** are critical in case the agent encounters issues, such as when it can’t understand the user’s request or when the user enters invalid data. * **Example**: If a user requests a product that is out of stock, the prompt can instruct the agent to provide alternatives or offer to **notify the customer when the product becomes available**. **Escalation Protocols** NexaVoxa allows you to build **escalation protocols** directly into the prompts. These are critical for ensuring that calls are efficiently transferred to human agents when the AI reaches its limits—either because of user frustration, failure to resolve a query, or a request for a more complex task. * **Example**: If the agent detects that the user has asked for something outside its capabilities or if the sentiment analysis flags frustration, the prompt can trigger an **immediate escalation to a live agent**. **Dynamic Adjustments Based on User Profile** Prompts can be **dynamically adjusted** based on the user’s profile or previous interactions. By pulling information from a CRM or the user's interaction history, NexaVoxa tailors responses to ensure a personalized experience for each user. * **Example**: If a returning customer is calling about a recent order, the agent might pull up their **order history** and customize the response based on previous purchases or inquiries. For a first-time customer, the agent could introduce the company and products more extensively. *** #### How Prompting Works in NexaVoxa NexaVoxa’s **drag-and-drop flow builder** lets you visually design complex conversational pathways, and **define prompts** for each step. These prompts are embedded within the call flow, determining the AI’s next action or response. The intuitive interface allows you to: * **Create multiple prompts** for different scenarios (sales, support, etc.) * **Insert conditional logic** into prompts to control behavior (e.g., repeat or escalate) * **Modify tone and language dynamically** based on user profiles, data, and intent * **Test and optimize prompts** using A/B testing to see which responses work best for specific customer segments. --- # Agent Instructions: Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter: ``` GET https://docs.nexavoxa.com/voice-intelligence-tech/custom-prompting.md?ask= ``` The question should be specific, self-contained, and written in natural language. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.