Skip to main content

​
Data Assets

Data Assets are the foundational building blocks of your AI workflows. They provide the structure and organization needed to develop, test, and optimize AI applications systematically.

Prompts

Instructions that guide AI model behavior with role-based messaging and schema validation.

Datasets

Input/output pairs for training, evaluation, and testing AI models with flexible data formats.

Metrics

Custom evaluation functions to measure model performance and optimize results.

​
Key Features

  • Schema Validation: Ensure data quality with input/output schema enforcement
  • Version Control: Track changes and iterations across all assets
  • Performance Monitoring: Real-time tracking of costs, speed, and accuracy
  • Multi-Provider Support: Work with OpenAI, Anthropic, and other leading AI providers

​
Getting Started

Data Assets work together to create a comprehensive AI development workflow:
  1. Prompts define how AI models should behave and what outputs to expect
  2. Datasets provide the input/output pairs needed for evaluation and testing
  3. Metrics measure performance and guide optimization decisions

Create Your First Prompt

Learn to build effective prompts with schema validation

Build a Dataset

Create and manage datasets for evaluation

Define Custom Metrics

Build evaluation functions to measure performance