Generator Agent

The Generator Agent is designed to automate structured and unstructured content creation. In the AI-Horizon platform, Generator Agents are specialized for discrete creative and synthesis tasks, ranging from generating synthetic training datasets to drafting professional content.

Capabilities

AI-Horizon comes with multiple pre-configured generators out of the box:

• Product Description Generator: Creates persuasive copy based on features, categories, keywords, and target audiences.
• Dataset Generator: Builds custom synthetic databases or lists (e.g., CSV rows for testing) given columns/labels and a high-level topic.
• LinkedIn Post Generator: Crafts engaging professional updates with customizable topics and tones.
• Logo/Image Generator: Leverages diffusion models to generate branding assets and user interface elements.

API Execution Pattern

Generators accept structured JSON parameters and return parsed results instantly. Developers can call these agents using REST APIs:

Example: Dataset Generation Request

curl -X POST http://localhost:5001/api/agents/dataset-generator/generate \
  -H "Content-Type: application/json" \
  -d '{
    "custom_labels": ["Ticket ID", "Customer Name", "Query", "Severity"],
    "rows": 5,
    "topic": "e-commerce support tickets"
  }'

Python SDK Integration

from aih import Generator
 
# Create a generator client
dataset_gen = Generator.dataset(api_key="your_aih_api_key")
 
# Generate synthetic tabular data
data = dataset_gen.generate(
    columns=["User", "Activity", "Duration"],
    rows=10,
    topic="workplace logs"
)
print(data.csv_format)