ConceptsWhat is mock data?

What is mock data?

Mock data is artificial data that corresponds to some broad criteria, but does not necessarily correlate to real-world trends based on some underlying subject dataset.

Mock data is distinct from synthetic data, which is artificial data that retains the characteristics of real-world data. However, a key feature of synthetic data is that it does not directly correspond to real-world entities like people, organizations, institutions, and others contained in the original dataset.

What is AI-generated mock data?

MOSTLY AI provides capabilities to generate mock data using the MOSTLY AI Assistant or via the MOSTLY AI Mock library.

Tabular mock data

Generative AI models

MOSTLY AI offers support for many generative AI models, including those from providers like OpenAI, Anthropic, Meta, Mistral, and others. For the entire library of available models, see the LiteLLM Providers documentation. You can also provide your own model hosted on Hugging Face.

Why AI-generated mock data?

AI-generated mock data allows you to quickly create realistic, customizable datasets without exposing sensitive information, making it ideal for testing, prototyping, and development.

With AI-generated mock data, you can:

  • Use LLMs to generate any tabular data to suit your needs.
  • Create entire datasets from scratch, including those with complex relationships.
  • Expand existing datasets to include new columns.
  • Enrich existing tables with new columns.
  • Use realistic and private data for testing and development.