Skip to main content

ā€‹
Overview

The AI Developer is an intelligent agent that transforms natural language descriptions into functional code. It helps developers generate business rules, JavaScript expressions, and test data without requiring manual coding. The agent can also edit existing code, provide explanations, and translate between programming languages, significantly accelerating the development process.

ā€‹
Key capabilities

The AI Developer offers two main categories of capabilities:

ā€‹
Generate new code

Generate Business Rule

Transforms natural language descriptions into code using the data model as context. Supports JavaScript, Python, and MVEL. Produces clean, functional business rule code based on text prompts.

Generate Business Rule from Document

Extracts business rule definitions from uploaded documents. Identifies relevant text on each page that describes business rules, summarizes and presents the information to users in a dropdown, and generates corresponding code based on the selected rule description.

Generate (un)Hide JavaScript Expression

Creates JavaScript expressions for conditional visibility of elements. Uses the data model as context for proper variable references.

Generate Computed JavaScript Expression

Produces JavaScript code for calculated or derived values. Uses the data model to ensure proper variable referencing.

Generate Custom Component

Generates a React Custom Component bundle (JSX, CSS, package.json, and sample input data) from a natural-language description. A sample input field lets you supply preview data so the generated component renders meaningfully on first run.

ā€‹
Edit existing code

Edit Existing Business Rule

Modifies existing code based on natural language instructions. Uses the data model as context for accurate updates.

Edit Custom Component

Iterates on an existing Custom Component bundle through the same conversation. The agent keeps the component context across turns and updates JSX, CSS, package.json, or sample input data as requested.

Fix / Explain Code

Fixes errors in business rules based on error logs. Provides clear explanations of what the code does.

Translate Code Between Languages

Converts business rules from one language to another. Maintains functionality while adapting to language-specific syntax.

ā€‹
Generate workflow script-node code

SaaS Ā·
Available on SaaS with FlowX.AI . This feature is live on managed (SaaS) deployments now. Self-hosted deployments receive it with the next LTS release.
In the workflow designer, a Script node can call AI Developer to generate or refactor its code from a natural-language prompt. The generated code streams into the nodeā€™s test area for a Keep or Discard review, with edits shown as a diff. It supports JavaScript and Python, and respects the same AI Developer permissions as the rest of the agent. To regenerate from scratch, clear the code first; configuration-parameter values are never sent to the model.

ā€‹
Project data model awareness

Available starting with FlowX.AI 5.9.2The AI Developer receives your projectā€™s data model as context when generating or editing business rules, hide/unhide expressions, and computed expressions.
Generated code references the data-model keys that already exist in your project. When the requested logic needs keys that are not in the data model yet, the agent proposes data-model changes alongside the code: new keys to create and existing project data types to instantiate. The proposed changes are presented for review, so you decide whether to add them to the data model before using the generated code.

ā€‹
How to use

The AI Developer is accessible through the FlowX.AI Platform interface. You can:
1

Select the desired capability

Choose the capability you need from the AI Developer section.
2

Provide instructions or upload documents

Provide natural language instructions or upload relevant documents.
3

Specify programming language preferences

Specify programming language preferences when applicable.
4

Review and implement the generated code

Review the generated code and implement it in your process.
For document-based generation, you can upload documents, view extracted rule descriptions in a dropdown, select the rule you want to generate, and receive the corresponding code.

ā€‹
How it works

The AI Developer architecture integrates large language models with the FlowX.AI Platform to convert natural language to code.

ā€‹
Best practices

ā€‹
āœ… DO the following

Be specific with your requirements: Clearly state what the business rule should accomplish, including conditions, actions, and expected outcomes.

Reference data model elements: Mention specific data fields or entities that the code should interact with to ensure proper context.

Specify the programming language: Always indicate whether you need JavaScript, Python, or another supported language for your code generation.

Provide edge cases: Mention special conditions or exceptions that the business rule should handle to ensure reliable code.

ā€‹
ā›” DONā€™T do these

Donā€™t be vague: Avoid ambiguous descriptions that can lead to incorrect implementation. Be precise about the logic you need.

Donā€™t reference external systems: The AI Developer has no knowledge of external systems not defined in the provided context.

Donā€™t skip defining variables: Make sure to define all variables or data fields that your business rule will use.

Donā€™t expect platform-specific knowledge: The agent doesnā€™t know about keys or information not defined in the data model or provided context.

Agent Builder

Build and configure custom AI agents for your business needs

AI Analyst

Generate process definitions and data models from natural language descriptions

AI Designer

Automatically generate UI components and screen layouts from process definitions
Last modified on July 13, 2026

Related topics

FlowX.AI 5.10.0 Release NotesFlowX.AI 5.8.0 Release NotesFlowX.AI 5.9.0 Release NotesAI Platform setupAI Gateway setup