Why Giskard?

Giskard is a holistic Testing platform for AI models to control all 3 types of AI risks: Quality, Security & Compliance.

It addresses the following challenges in AI testing:

  • Edge cases in AI are domain-specific and often seemingly infinite.

  • The AI development process is an experimental, trial-and-error process where quality KPIs are multi-dimensional.

  • Generative AI introduces new security vulnerabilities which requires constant vigilance and continuous red-teaming.

Giskard provides a platform for testing all AI models, from tabular ML to LLMs. This enables AI teams to:

  1. Reduce AI risks by enhancing the test coverage on quality & security dimensions.

  2. Save time by automating testing, evaluation and debugging processes.

Giskard Library (open-source)

An open-source library to detect hallucinations and security issues to turn them into test suites that you can automatically execute.

Testing Machine Learning applications can be tedious. Since AI models depend on data, quality testing scenarios depend on domain specificities and are often infinite. Besides, detecting security vulnerabilities on LLM applications requires specialized knowledge that most AI teams don’t possess.

To help you solve these challenges, Giskard library helps to:

  • Detect hallucinations and security issues automatically: The giskard RAGET and SCAN automatically identify vulnerabilities such as performance bias, hallucination, prompt injection, data leakage, spurious correlation, overconfidence, etc.

  • Instantaneously generate domain-specific tests: giskard automatically generates relevant, customizable tests based on the vulnerabilities detected in the scan.

    ../_images/test_suite_scan_llm.png

  • Integrate and automate testing of AI models in CI/CD pipelines by leveraging native giskard integrations.

    ../_images/gh_discussion.png

Get started now with our quickstart notebooks! ⚡️

Want to take Giskard’s features to the next level? Discover the LLM Hub below!

LLM Evaluation Hub (for entreprises)

The LLM Hub is an enterprise solution offering a broader range of features such as a:

  • Continuous testing: Detect new hallucinations and security issues during production.

  • Annotation studio: Easily write the right requirements for your LLM as a judge setup.

  • Alerting: Get alerted with regular vulnerability reports by emails during production.

  • Red-teaming playground: Collaboratively craft new test cases with high domain specificity.

For a complete overview of LLM Hub’s features, check the documentation of the LLM Hub or directly contact us.