Giskard - Testing platform for AI models

Created

Apr 8, 2024 7:45 AM

AI keywords

AI TestingBias DetectionPerformance EvaluationSecurity AssessmentComplianceOpen-source Testing FrameworkPython LibraryEnterprise HubCollaborative AI QualityML ModelsLLMsEU AI ActData ScientistsML EngineersAI Governance Officers

AI summary

Giskard is a testing platform for AI models that helps protect against biases, performance issues, and security vulnerabilities. It offers an open-source testing framework, a Python library, and an enterprise hub for collaborative AI quality, security, and compliance. Giskard automates the detection of performance, bias, and security issues in AI models, saving time on manual testing and custom evaluation reports. It also ensures compliance with the EU AI Act. The platform is suitable for data scientists, ML engineers, and AI governance officers who prioritize performance, security, and safety in AI models.

Text

Giskard

URL

https://www.giskard.ai/

Protect your company against biases, performance issues & security vulnerabilities in AI models.

From tabular models to LLMs

Listed by GartnerAI Trust, Risk and Security

# Get started

pip install giskard[llm]

You can copy code here

Giskard - Open-source testing framework for LLMs & ML models | Product Hunt

Trusted by leading AI teams

Why?

AI pipelines are broken

AI risks, including quality, security & compliance, are not properly addressed by current MLOps tools.

AI teams spend weeks manually creating test cases, writing compliance reports, and enduring endless review meetings.

AI quality, security & compliance practices are siloed and inconsistent across projects & teams,

Non-compliance to the EU AI Act can cost your company up to 3% of global revenue.

Enter Giskard: AI Testing at scale

Automatically detect performance, bias & security issues in AI models.

Stop wasting time on manual testing and writing custom evaluation reports.

Unify AI Testing practices: use standard methodologies for optimal model deployment.

Ensure compliance with the EU AI Act, eliminating risks of fines of 3% of your global revenue.

Mitigate AI Risks with our holistic platformfor AI Quality, Security & Compliance

Giskard Library

Open-Source Open-source Python library to identify & control risks in ML models and LLMs automatically, with a complete test coverage of performance, ethics & security metrics.

Giskard HubEnterprise Hub for teams to collaborate on top of the open-source library, with compliance dashboards, debugging, human feedback, explainability and secure access controls.

Open-source & easy to integrate

In a few lines of code, identify vulnerabilities that may affect the performance, fairness & security of your model.

Directly in your Python notebook or Integrated Development Environment (IDE).

import giskard

qa_chain = RetrievalQA.from_llm(...)

model = giskard.Model(

qa_chain,

model_type="text_generation",

name="My QA bot",

description="An AI assistant that...",

feature_names=["question"],

)

giskard.scan(model)

Giskard Hub

Collaborative AI Quality, Security & Compliance

Entreprise platform to test, debug & explain your AI models collaboratively.

Who is it for?

Data scientists

ML Engineers

AI Governance officers

You work on business-critical AI applications.

You spend a lot of time to evaluate AI models.

You want to work with the best Open-source tools.

You’re preparing your company for compliance with the EU AI Act and other AI regulations.

You have high standards of performance, security & safety in AI models.

"Giskard really speeds up input gatherings and collaboration between data scientists and business stakeholders!"

Head of Data

Emeric Trossat

Join the community

Welcome to an inclusive community focused on AI Quality, Security & Compliance! Join us to share best practices, create new tests, and shape the future of AI standards together.

Discord

All those interested in AI Quality, Security & Compliance are welcome!

All resources

Knowledge articles, tutorials and latest news on AI Quality, Security & Compliance

Our new course in collaboration with DeepLearningAI team provides training on red teaming techniques for Large Language Model (LLM) and chatbot applications. Through hands-on attacks using prompt injections, you'll learn how to identify vulnerabilities and security failures in LLM systems.Red Teaming LLM Applications course

Introducing our LLM Red Teaming service, designed to enhance the safety and security of your LLM applications. Discover how our team of ML Researchers uses red teaming techniques to identify and address LLM vulnerabilities. Our new service focuses on mitigating risks like misinformation and data leaks by developing comprehensive threat models.Giskard's LLM Red Teaming

Learn how to effectively monitor and manage data drift in machine learning models to maintain accuracy and reliability. This article provides a concise overview of the types of data drift, detection techniques, and strategies for maintaining model performance amidst changing data. It provides data scientists with practical insights into setting up, monitoring, and adjusting models to address data drift, emphasising the importance of ongoing model evaluation and adaptation.Data Drift Monitoring with Giskard

Ready. Set. Test!Get started today

Get started

Stay updated with

the Giskard Newsletter

Giskard is the first holistic platform to ensure the quality, security & compliance of all AI models, built for the EU AI Act.

Cookies settings