Building AI Agents | Part 5: Defining Properties & Data Types

Welcome to Part 5 of our series, where we explore the essential foundation of any high-performing AI Agent: Properties. Learn how to define the fields and data types your agent will extract, classify, and calculate, ensuring your outputs are structured, accurate, and ready for any business workflow. Using a real-world Purchase Order Processing agent from a logistics workflow, we provide a practical, step-by-step guide to every major property type available in V7 Go. Discover when and why to use each one, including: - Text & Number: For flexible and numerical data with built-in AI citations for auditability. - Single & Multi Select: For classifying data against predefined lists and building powerful routing logic. - Collections: The perfect tool for handling tabular data like line items in an invoice or purchase order. - JSON: For efficiently extracting multiple related key-value pairs in a single, structured output. - Reference: An advanced property for creating powerful, interconnected workflows by linking multiple agents together. We also share best practices for choosing the right property type based on your specific needs, helping you build agents that are scalable, reliable, and perfectly aligned with your business goals. 0:00 - Introduction to Agent Properties & Data Types 0:13 - Use Case: Purchase Order Processing in Logistics 0:26 - What is a Property? (Like Columns in a Spreadsheet) 0:33 - The 3 Core Components of a Property 0:54 - Property Type: Text (for flexible outputs & summaries) 1:11 - Property Type: Number (for structured financial data) 1:32 - Property Type: Single Select (for classification & routing) 1:55 - Property Type: Multi Select (for applying multiple tags) 2:13 - Property Type: Collections (for handling tables & line items) 2:50 - Property Type: JSON (for nested key-value pairs) 3:11 - Property Type: Reference (for linking agents) 3:35 - Best Practices for Choosing the Right Property Type Find out more on how V7 Go's AI Agents can help your business: https://www.v7labs.com/go/agents Learn more about V7 - https://www.v7labs.com Follow us on LinkedIn: https://www.linkedin.com/company/v7labs/ Follow us on X (formerly Twitter): https://twitter.com/V7Labs #AIagents #LogisticsTech #PurchaseOrder #DataTypes #V7Go #WorkAutomation #GenerativeAI #StructuredData #AgenticAI #DocumentProcessing