Feature Availability

Vtiger Editions: One Growth | One Professional | One Enterprise | One AI 

Introduction 

Prediction AI Designer helps you predict future outcomes using your past CRM data. It allows Admins and authorized users to create predictions for different modules so that teams can take action early instead of reacting later.

With this AI-powered and no-code feature, you can easily forecast results such as deal outcomes, task completion, or case resolution directly inside Vtiger CRM.

Types of Predictions

Prediction AI Designer supports the following types:

• Data Classification – Predicts items to categories.
  • Example: An email is classified as spam or not spam.
• Data Prediction – Predicts continuous values such as dates or numbers.
  • Example: Predicting next month’s sales.
• Data Scoring – Assigns a numeric likelihood or risk score.
  Example: A customer’s credit score of 750 or a lead’s likelihood to convert.

Benefits of Prediction AI Designer

The benefits of using Predictive AI Designer are:

• Helps you make better decisions using AI-based insights.
• Saves time by automatically analyzing historical data.
• Predicts important outcomes like renewals, delays, or churn.
• Easy to use with a no-code setup.
• Uses existing CRM data without manual effort.

Enabling Predictive AI

To use Predictive AI Designer, you must install the Calculus AI add-on.

Follow these steps to install the Calculus AI: 

1. Log in to the CRM.
2. Click the User Menu.
3. Select Settings. The Settings page opens. 
4. Go to Add-ons. The Add-ons page opens. 
5. Search for Calculus. 
6. Click Try for 30 days to install it.

Prerequisites for Creating a Prediction

Before creating a prediction, make sure:

• You have at least 50 historical records. More data gives better results.
• Historical records contain the actual value of the field you want to predict, either directly or through a calculation.

Adding a Predictive AI

Follow the steps below to add a predictive AI: 

• Step 1: Enter Prediction Details
• Step 2: Configure Parameters
• Step 3: Select Data Training
• Step 4: Set Frequency 

Step 1: Enter Prediction Details

Follow these steps to enter the Prediction details:

1. Log in to the CRM.
2. Click the main Menu.
3. Go to Platforms > Predictive AI Designer. The Predictive AI Designer page opens. 
4. Click + Add New Models. The Add Prediction opens. 
5. Enter or select the following information:
   1. Prediction Name: Enter a meaningful name for the prediction.
   2. Select Module: Choose the module for which you want to create the prediction.
   3. Description: Briefly describe the purpose of the prediction.
   4. Category: Select one of the following:
      1. Data Classification
      2. Data Prediction
      3. Data Scoring
   5. Default Algorithm: An algorithm is selected automatically. You can keep the default setting.
6. Click Next.

Step 2: Configure Parameters

You choose the fields that influence the prediction and define what value needs to be predicted.

Follow these steps to select the Parameters:

1. Select Influencing Fields
   1. Select the fields that may affect the predicted value. These fields help the AI identify patterns in historical data.
   2. Use the Search fields option to quickly find fields.
2. Is the Value to Predict Available in Historical Records? Select one option:
   1. Yes – Choose this if the value to be predicted already exists in a field.
      1. Select the Prediction Field from the list.
   2. No – Choose this if the value is not directly available.
      1. You must configure a transformation or formula to compute the value.
      2. Adding at least one Before Training Transformation is mandatory. 
      3. Note: The prediction model learns from historical records.
   3. Data Transformation: You can apply transformations to prepare or adjust data:
      1. Before Training Transformation
      2. Before Prediction Transformation
      3. After Prediction Transformation
   4. Click Next

Step 3: Select Training Data

In this step, you choose which CRM records are used to train the prediction model.

Follow these steps to select training data:

1. Select Use data from CRM
   1. By default, all records that contain values are used for training.
2. Filter Training Data: You can narrow down training data using conditions:
   1. All Conditions: All conditions must be met.
   2. Any Conditions: At least one condition must be met.
3. Use View XML Script to view the condition logic in XML format.
4. Click Next.

Step 4: Set Frequency

Configure how often the model and prediction scores are updated.

Follow these steps to set frequency: 

1. Update Model Frequency
   1. Once a week: The model is retrained automatically every week.
   2. Never: The model is updated only when an Admin manually re-trains it. 
2. Update Prediction Score Frequency
   1. When the record is updated: Prediction scores refresh whenever a record changes. 
3. Click Save.

Reference 

• Calculus AI
• AI Agent Builder
• Vtiger Chatflows
• Calculus AI: Your Smart Assistant with AI Technology
• Natural Language Query
• Calculus GPT
• Deal Health
• Best Time to Contact
• Contact Sentiment
• Deal Recommendations
• Conversation Insight