About Giskard
Giskard is an open-source ML testing framework and commercial platform for detecting vulnerabilities and testing AI models (from tabular models to LLMs). It provides automated scans, test-generation, a collaborative hub for domain experts, and tools for red-teaming LLMs and building test suites to prevent regressions.
Key Features
- Automated vulnerability scanning — detects biases, leakage, hallucinations, overconfidence and other issues
- Test suite generation & execution — generate and run unit-style tests and domain-specific scenarios
- Open-source SDK + Hub — Python SDK for devs and an enterprise Hub for collaboration, annotation and red-teaming
- LLM-focused testing — tools and benchmarks for hallucination, factuality and security testing
Use Cases & Best For
Developers and QA teams building test suites and automated scans for ML and LLM systems
Product and domain experts who need no-code/collaborative tooling to define business tests and red-team models
About Model Evaluation
Test and evaluate AI models