AI Prompts

AI-driven capabilities, powered by configurable AI Prompts, enable organizations to analyze customer sentiment, control how AI interprets and responds to interactions, and visually represent customer emotions across chats and CRM records. These prompts act as structured instructions that guide the AI on how to think, analyze, and respond across different use cases such as sentiment detection, text refinement, and contextual assistance.

By transforming raw conversation data into clear, actionable insights, this approach provides agents with immediate visibility into a customer’s emotional state, allowing them to respond with the right tone, deliver more personalized interactions, and manage conversations more effectively to enhance the overall customer experience.

With AI Prompts, organizations can:

• Customize AI behavior based on business needs

• Allow agents to define instructions on how AI functionalities should work

• Reuse predefined prompts for faster interactions

After license activation, AI prompts are available by default, and users can create additional prompts for quick and repeated use.

AI Prompt Configuration

Each field in AI Prompt configuration defines how the prompt is identified, prioritized, and applied within the system.

Title: used to assign a clear, user-friendly name to the prompt. Unique Name: automatically generated by the system based on the title and serves as an internal identifier. Order: controls how prompts are arranged and displayed in the user interface. System Instruction: contains predefined logic provided by the system. These instructions are available only in the default prompts provided by Inogic and come pre-configured and optimized after license activation. User Instruction: defines custom guidelines that direct how the AI should behave and respond. When both system and user instructions are present, user instructions take the highest priority and override system instructions. Favourite: allows users to mark prompts for quick access within the interface. Prompt Target: determines where the prompt will be applied, such as in indicators, text rephrasing, or within the AI panel. AI Feature: specifies the AI capability being used. AI Model: allows selection between the default model or a custom-configured model. Schema Details: define the technical structure and mapping required for the prompt to function correctly.

Favourite Prompts

• Favourite prompts appear in the InoWiz Panel on the chat interface.

• Based on order, top 5 marked favourite will be shown in the interface.

• Helps agents access AI suggestions instantly.

Prompt Targets (Where the Prompt Appears)

You can control where the AI prompt is applied based on its purpose and functionality within the system.

Text Rephrase is used to improve the tone, grammar, and overall phrasing of messages. It appears in the text area while the user is typing, helping refine responses before sending.

Indicator is used to generate sentiment or emotion indicators. It is visible on the chat header as well as on related CRM records, allowing quick identification of customer sentiment.

InoWiz Panel is available within the chat interface, where the prompt provides contextual AI assistance and insights during conversations.

What Are AI Indicators?

AI Indicators are visual representations of customer sentiment generated by AI Prompts. They help agents quickly identify customer emotions without reading the entire conversation.

Indicators can include:

• Text labels (e.g., Happy, Neutral, Sad)

• Colors for quick visual recognition

• Emojis to make sentiment easier to understand briefly.

These indicators appear directly in chats and on CRM records, making emotional context instantly visible.

Indicators can be customized beyond WhatsApp4Dynamics default AI sentiments.

AI Indicator Configuration

Indicator Settings define how sentiment is visually represented.

By default, the system provides 3 sentiments:

• Positive

• Neutral

• Negative

Accessing the Default Sentiment Prompt

To view and manage the default sentiment configuration, navigate to:

AI Prompts → Sentiment Analysis

• The default sentiment analysis prompt is pre-configured and ready to use. However, it is fully flexible and can be modified based on your business requirements. You can update the instructions, adjust how sentiments are interpreted, or customize how they are displayed.

• If the default sentiment setup does not align with your needs, you also have the option to either deactivate it or delete it entirely. This allows you to create and use your own custom sentiment prompts for more tailored behavior and representation.

Steps to Create a New Indicator Prompt:

• Navigate to AI Prompts • Click on New • Fill in all the required details

• Add Row option allows you to create additional indicators.

• Select the required Entity for which the sentiment should be applied.

You can customize:

• Indicator text

• Emoji

• Color

• Description

• Number of sentiment levels

Click Save.

Entity-Specific Indicators

Indicators can be entity-specific, allowing sentiment to be stored and displayed per entity.

Only entities with record types configured will be displayed here.

E.g.

• Account

• Contact

• Lead

If sentiment is created for the Chat entity, it will be visible on the Chat entity as well as on the associated customer record (Contact, Account, Lead).

Entity-Specific Behavior (Example: Contact)

If configured for Contact entity:

• The text field will be on the Chat entity, and the user must add that entity in the Mapping Field on the prompt.

• Sentiment appears:

On the Contact record

• If the Contact is set as the Customer:

Sentiment appears both on Chat entity and Contact entity

Mapping Field

• To store sentiment data, you must create a text field on the Chat entity where the sentiment value will be saved.