GreenData has integrated an AI agent into its platform to assist with algorithm development. The tool allows users to generate code by describing business logic in natural language, directly within the algorithm editor. The agent operates within a chat interface, creating code based on platform-specific rules and context.
The AI agent supports both the creation of new algorithms and the modification of existing ones. It can automatically attempt to fix compilation errors or broken references. Users can provide additional context by linking objects, methodologies, or specific code snippets to their prompts to refine the output.
The platform features a comparison mode where AI-suggested changes are highlighted, allowing users to accept or reject modifications block by block. A version history is available to revert to previous states. The tool also supports collaborative work, as team members can view shared chat logs and version history.
According to Alexander Perevalov, head of the AI development group at GreenData, the agent is designed to account for application data models, including object types, attributes, and relationships. While the AI handles the generation process, human specialists retain control over verifying the logic and its impact on business operations.
