Support docs

Prompt Guide: How do I customize my bots in terms of persona, reading level, length, and so forth?

Find out how you can easily customize each bot for their specific audience with QueryPal.

Introduction

QueryPal offers deep bot customization over a variety of dimensions, based on the audience and desired outcome for that bot's interactions. To make any of the changes described below, go to the bot config page (at https://admin.querypal.com/bot-configurations), select the specific bot, open up the 'Advanced' section, and you'll find the instructions field below. 

bot-instructions

You can add multiple different customizations into the same instruction field; just make sure to click 'Save' after you're done. 

Here are just some of the customizations we've seen customers leverage with QueryPal. If you'd like to learn more or get assistance with your customization, don't hesitate to email us at support@querypal.com

Language Localization and Regional Adaptation

Description: Customize the bot's language to match your target audience's regional preferences.

Example Prompt: "Use British English spelling and vocabulary. Incorporate common British idioms where appropriate."

How it modulates:

User: "What colour is the new product?"

Bot: "The new product comes in a lovely shade of blue, one of several colours we offer."

Audience-Specific Interpretation

Description: Tailor the bot's understanding and responses to a specific professional context.

Example Prompt: "Interpret questions as coming from Platform Engineers at a large-scale SaaS company."

How it modulates:

User: "How do we handle scaling?"

Bot: "For scaling our SaaS platform, we employ a combination of horizontal and vertical scaling strategies. Our Kubernetes clusters auto-scale based on CPU and memory utilization, while our database layer uses read replicas and sharding for improved performance."

Topic Management and Redirection

Description: Define the scope of discussions and provide appropriate redirects for out-of-scope queries.

Example Prompt: "Focus on product features and technical support. Avoid discussing pricing or competitors. For out-of-scope queries, politely redirect to the appropriate department."

How it modulates:

User: "How does your pricing compare to Competitor X?"

Bot: "I apologize, but I'm not able to discuss pricing or make comparisons with other companies. For the most up-to-date pricing information, please contact our sales team at sales@company.com."

Disclaimer Configuration

Description: Set up necessary disclaimers for sensitive or regulated topics.

Example Prompt: "Include a data privacy disclaimer when discussing user data or analytics."

How it modulates:

User: "How do you handle user data?"

Bot: "Before I answer, please note: This information is general and may not cover all aspects of our data handling practices. For complete details, refer to our privacy policy.

We prioritize user data protection. All data is encrypted in transit and at rest, and we follow strict access control protocols..."

Keyword-Triggered Responses

Description: Create specific responses for important keywords or phrases.

Example Prompt: "When 'system outage' in mentioned, apologize and provide emergency contact information (phone: 1-800-555-1234 and email: urgent@company.com)."

How it modulates:

User: "Is there a system outage?"

Bot: "I'm sorry to hear you might be experiencing issues. For immediate assistance with potential system outages, please contact our 24/7 emergency support line at 1-800-555-1234 or email urgent@company.com."

Guided Conversation Flows

Description: Create structured conversations for specific scenarios.

Example Prompt: "For new users, guide them through a 3-step onboarding process: account setup, feature overview, and first task completion."

How it modulates:

User: "I'm new here. How do I get started?"

Bot: "Welcome! I'd be happy to guide you through our onboarding process. Let's start with setting up your account.

Step 1: Account Setup First, I'll guide you through choosing a secure password and setting up two-factor authentication.

After that, we'll move on to a quick overview of our main features and help you complete your first task. Ready to begin?"

Answer Length and Detail Control

Description: Adjust the verbosity and depth of the bot's responses.

Example Prompt: "Provide concise answers (2-3 sentences) for general queries. For technical questions, offer more detailed explanations (4-6 sentences)."

How it modulates:

User: "What is cloud computing?"

Bot: "Cloud computing is the delivery of computing services over the internet. It allows businesses to access resources like storage and processing power without maintaining physical infrastructure."

User: "How does containerization work in cloud environments?"

Bot: "Containerization in cloud environments involves packaging applications and their dependencies into isolated units called containers. These containers are lightweight, portable, and can run consistently across different computing environments. In cloud settings, container orchestration platforms like Kubernetes manage the deployment, scaling, and operation of these containers. This approach enhances application portability, improves resource utilization, and enables more efficient scaling and management of cloud-based applications."

 

Answer Structure Formatting

Description: Define how information should be presented in responses.

Example Prompt: "Use bullet points for listing features or steps. Use paragraphs for explanations and conceptual information."

How it modulates:

User: "What are the key features of your product?"

Bot: "Our product's key features include:

  • Real-time data analytics
  • Customizable dashboards
  • Automated reporting
  • Integration with major CRM systems
  • Mobile app for on-the-go access"

User: "How does your data analytics work?"

Bot: "Our data analytics system processes information in real-time, allowing for immediate insights. It uses advanced machine learning algorithms to identify patterns and trends in your data. This enables you to make data-driven decisions quickly and efficiently, giving you a competitive edge in your market."

Reading Level Adjustment

Description: Tailor the language complexity to match your audience's expertise.

Example Prompt: "Use technical language appropriate for IT professionals with a focus on clarity."

How it modulates:

User: "How does your system handle data replication?"

Bot: "Our system employs asynchronous multi-master replication for data consistency across distributed nodes. We utilize a combination of conflict-free replicated data types (CRDTs) and vector clocks to manage concurrent updates and ensure eventual consistency. This approach allows for high availability and partition tolerance in accordance with CAP theorem principles."

Tone and Persona Customization

Description: Define the bot's communication style and personality.

Example Prompt: "Maintain a professional yet friendly tone. Use polite language and occasional light humor where appropriate. Avoid emojis."

How it modulates:

User: "This is frustrating. Can you help?"

Bot: "I completely understand your frustration, and I'm here to help. Let's work together to resolve this issue. Could you please provide more details about what's troubling you?"

 

By implementing these prompts, you can create a highly customized and effective bot that aligns perfectly with your brand voice and meets the specific needs of your audience.