overview

The aiProcessing function performs flexible AI-powered data processing using custom prompts. This function gives you complete control over how AI analyzes, transforms, or generates content based on your specific requirements.

Executes custom AI processing tasks using natural language prompts. Unlike the structured extraction of aiExtractVariables, this function is designed for open-ended AI tasks like content generation, analysis, summarization, classification, or any custom processing logic you can describe in a prompt.

Key Differentiator: Unlike aiExtractVariables which extracts specific structured data, aiProcessing allows open-ended AI tasks that you can describe in natural language prompts.

Key Capabilities

Custom Prompt Processing

Execute any AI task you can describe in natural language

Flexible Output Formats

Return results as plain text or structured JSON

Image Analysis

Can extract information from images when provided with image buffers

Content Generation

Create new content based on input data and instructions

Data Transformation

Convert data between formats or apply complex business logic

Parameter	Type	Required	Description
prompt	string		Natural language instructions for the AI to follow
output_type	string	Optional	Format of response: "text" (default) or "json"
image_buffer	buffer	Optional	Image data to analyze alongside text
image_mimetype	string	Optional	MIME type of the image (e.g., "image/png")

Parameter

Type

Required

Description

prompt

string

Natural language instructions for the AI to follow

output_type

string

Optional

Format of response: "text" (default) or "json"

image_buffer

buffer

Optional

Image data to analyze alongside text

image_mimetype

string

Optional

MIME type of the image (e.g., "image/png")

Usage Examples

Simple Text Processing Example

Basic conversation summarization using aiProcessing:

Conversation Summarization

JSON Output Example

Analyzing customer sentiment and returning structured data:

Customer Sentiment Analysis

Image Processing Example

Analyzing document images alongside text prompts:

Document Image Analysis

Data Transformation Example

Converting data formats for API integration:

API Format Conversion

Advanced Content Generation Example

Generating professional email responses with specific requirements:

Email Response Generation

Using Previous Step Data

Creating comprehensive reports by combining data from multiple workflow steps:

Business Report Creation

JSON Output Note: When using output_type: "json", you must include a request in the prompt to respond as a JSON Object (e.g. "Provide your response as a JSON Object").

Best Practices

Be Specific in Prompts

Clear, detailed instructions produce better results. Specify exactly what you want the AI to do and what format you expect.

Provide clear, detailed instructions about what you want the AI to accomplish
Specify exactly what format you expect in the output
Include examples in your prompts when they help clarify requirements

Use Output Type Appropriately

Choose "json" when you need structured data that other steps can easily reference, and "text" for human-readable content.

Choose "json" when you need structured data for other steps to reference
Choose "text" for human-readable content and final responses
When using "json", explicitly request JSON Object format in your prompt

Include Context

Reference previous step outputs and relevant variables to give the AI complete context for processing.

Reference previous step outputs using variable interpolation
Include relevant variables and context from your workflow
Provide complete context so the AI can make informed decisions

Handle Variable Data

When working with data from previous steps, ensure your prompts can handle variations in data structure and content.

Ensure your prompts can handle variations in data structure and content
Structure your requests to work with different input scenarios
Consider token limits when working with very long prompts or large amounts of input data

aiProcessing

Table of Contents

overview

Key Capabilities

Custom Prompt Processing

Flexible Output Formats

Image Analysis

Content Generation

Data Transformation

input_parameters

Expected Outputs

Text Output (Default)

Text Response

JSON Output

JSON Response

Error Response

Error Output

Usage Examples

Simple Text Processing Example

Conversation Summarization

JSON Output Example

Customer Sentiment Analysis

Image Processing Example

Document Image Analysis

Data Transformation Example

API Format Conversion

Advanced Content Generation Example

Email Response Generation

Using Previous Step Data

Business Report Creation

Best Practices

Be Specific in Prompts

Use Output Type Appropriately

Include Context

Handle Variable Data

Related Functions