Browse Tasks

Select category

Select category

Select category

Select category

Select Task

Select Task

Select Task

Select Task

Select Tasks

Select Industry

Agent Playbook

Browse agentic AI tasks that instruct LLMs to carry out specific jobs.
These can be integrated into existing workflows in your apps and tools.

Claims Policy Verification

Claims Triage

Claims Fraud Analysis

Customer Query Categorisation

Policy Comparison

Damage Assessment from Images

Insurance Needs Analysis

Policy Renewal Recommendation

Claims Status Update

Compliance Documentation Check

Claims Documentation Validation

Query Triage and Escalation

E-commerce Manager

Product Categorisation

This agentic AI task categorises e-commerce products - vital for content management agents. The task configuration outlines how to:

  • Analyse product information for category alignment

  • Assign appropriate taxonomy classifications

  • Generate relevant product tags and attributes

  • Validate category placement accuracy

  • Ensure taxonomy consistency standards


As a modular task, it integrates with inventory and PIM systems, enabling AI agents to execute systematic product categorisation within existing e-commerce workflows. The task specification includes outputs for category assignment, attribute generation, and taxonomy validation, adaptable across various marketplace structures and product types.

Example outputs:

{
  "main_category": "Electronics",
  "subcategory": "Smart Home Devices",
  "suggested_tags": [
    "smart thermostat",
    "energy-saving",
    "Wi-Fi enabled",
    "mobile app control",
    "voice assistant compatible",
    "HVAC",
    "home automation"
  ],
  "confidence_level": "High",
  "alternative_category": "Home Improvement",
  "categorisation_notes": "The product is primarily an electronic device for smart home control, but it also relates to home improvement due to its HVAC functionality. The main category is chosen as Electronics due to its smart features and connectivity options.",
  "requires_human_review": false