Autopilot Batch

Autopilot Batch, part of the Autopilot Suite, enables you to run AI extractions across a large set of Salesforce records on-demand. Unlike the other Autopilot tools, Batch jobs are manually triggered and run against a large set of Salesforce records that match your defined criteria.

How Autopilot Batch Works

Autopilot Batch finds all records matching your filter criteria, runs your AI prompt against the historical conversational data for each one, and then updates a field in Salesforce.

• Run Batch
• Find Records
• For Each...
• Analyze History
• Run AI Prompt
• Update Field

Use Cases for Autopilot Batch

• Quarterly Health Checks: Run a batch job on all “Active” accounts to identify potential churn risks by summarizing recent support-related call snippets.
• Data Enrichment: For all leads created last month from a specific campaign, run a batch to extract and populate the “Primary Pain Point” field.
• Competitive Analysis: Identify all opportunities where a specific competitor is mentioned and extract the context for a competitive intelligence report.

Step-by-Step Configuration

Step 1: Access the Autopilot Batch Section

1. Log in to your Momentum admin dashboard.
2. Navigate to the Autopilot section and select the Batch tab.
3. Click “New Batch Extraction” to begin.

Step 2: Define the Target Records

Create a filter to select the Salesforce records you want to process. The batch will run on all records that currently match these conditions when you trigger it.

• Object: Select the Salesforce object (e.g., Opportunity, Account, Lead, Event).

Autopilot Batch does not currently support Contact objects. Contact field extractions are available in Autopilot Classic.

• Field Conditions: Add one or more conditions to build your list.
  • Field: e.g., Status, CreatedDate, ARR__c.
  • Operator: e.g., equals, greater than, is not null.
  • Value: The value to match.

Example Filter:

Object: Account
Conditions:
  - Status: equals "Active"
  - ARR__c: greater than 50000
  - LastActivityDate: within "Last 90 Days"

Step 3: Set the Time Window

Define the period of historical conversational data that the AI should analyze for each record in the batch.

Step 4: Write Your AI Prompt

This prompt will be executed for each record in the batch, using its associated historical conversational data.

Example Prompts:

• For an Account Health Check:

"Review all calls and emails for this account from the last 6 months. Summarize any mentions of product dissatisfaction, support issues, or churn risk. If none are found, state 'No risks identified'."

• For Lead Data Enrichment:

"Analyze the initial discovery call for this lead. What is their primary business challenge? Respond with a single, concise sentence."

Step 5: Configure Save Behavior

Choose how the extracted data is saved to Salesforce. Given the bulk nature of batch jobs, Write if Empty or Automatic Write are common choices after testing.

• Confirm to Write
• Write if Empty
• Automatic Write

Handling Records with No Context: You can configure what happens when no relevant historical conversational data is found for a record:

• Skip: Nothing will be written to Salesforce (default behavior)
• Write Reason: Write a message explaining why no data could be extracted
• Custom Text: Write custom text that you specify

This ensures fewer missed write-backs in edge cases and maintains data consistency even when context is missing.

This setting still respects your “Write and apply the suggestions” configuration. Confirm to Write, Write if Empty, and Automatic Write settings will still apply.

Autopilot automatically truncates extracted values to fit the target field’s maximum length in Salesforce.

Step 6: Preview and Test

Before running the full batch, use the Preview feature to validate your configuration:

1. Click the “Preview” button in the editor wizard
2. Preview tests your prompt against a sample of matching records
3. Review the AI-generated results to ensure accuracy
4. Refine your prompt based on preview feedback
5. Iterate quickly until you’re satisfied with the output quality

The Preview feature enables rapid iteration and validation of prompts before committing to a full batch run.

Step 7: Create and Run the Batch Job

Once you’ve validated your configuration with Preview:

• Create Batch: Set up the batch job with your configured settings
• Review Tasks: Check the list of records that will be processed from the autopilot batch dashboard
• Run Batch: Execute the extraction on all records that match your criteria
• Monitor Progress: Track the batch job status from the autopilot batch dashboard
• Review Results: Check completed extractions and any errors in the job report