Giskard AI
Giskard is the leading open-source AI testing platform for quality, security, and compliance, specializing in automated red teaming for LLMs and AI agents.
At a Glance
- Enterprise AI Teams
- Financial Services
- Automotive
- Healthcare
- +1 more
AI Tools by Giskard AI
(1)
Giskard
LLM Security Testing Platform
Discussions
No discussions yet
Be the first to start a discussion about Giskard AI
Latest News
Giskard announces Phare, a new LLM Evaluation Benchmark at Paris AI Summit
Study: Asking chatbots for short answers can increase hallucinations
Announcing Giskard OSS v3: Rebuilt for the Agentic Era
Giskard selected for second GitHub Accelerator cohort
Products & Services
A Python library to automatically detect vulnerabilities in machine learning models and LLMs, including hallucinations, bias, and performance issues.
A collaborative platform for AI teams to manage, monitor, and automate model testing workflows, offering both SaaS and on-premise deployments.
A specialized module for assessing and improving computer vision models, such as image classification and object detection.
An independent LLM evaluation benchmark designed to measure the trustworthiness and safety of large language models.
Market Position
Positioned as the first automated red teaming platform specifically for AI agents, leveraging an open-source framework to drive community-led security standards.
Leadership
Founders
Alex Combessie
Co-founder & Co-CEO. Previously Machine Learning Engineer at Dataiku and Strategy Consultant at Capgemini Consulting. Holds an MSc from HEC Paris and Telecom Paris.
Jean-Marie John-Mathews
Co-founder & Co-CEO. Previously Data Scientist at Thales and Capgemini. PhD in AI Ethics and AI Safety researcher. University lecturer at Sciences Po and PSL.
Andrei Avtomonov
Co-founder and former CTO. Previously Senior Software Engineer at Dataiku and Software Engineer at Amazon.
Executive Team
Alex Combessie
Co-Founder & Co-CEO
Ex-Dataiku, Ex-Capgemini, AI expert.
Jean-Marie John-Mathews
Co-Founder & Co-CEO
PhD in AI Ethics, Ex-Thales, AI Safety researcher.
Board of Directors
Founding Story
Alex Combessie and Jean-Marie John-Mathews met at Capgemini in 2015. After working as engineers at Dataiku, they joined forces with Andrei Avtomonov to build Giskard, aiming to bring the same rigor to AI testing as software engineering.
Business Model
Revenue Model
Open-core model with subscriptions for the Enterprise Hub (SaaS/On-prem) and professional support.
Pricing Tiers
Python library for automated testing, scanning, and red teaming of ML/LLM models.
Individual and small team plans for managing AI quality in the cloud.
Full collaborative suite with advanced security, on-premise deployment, and priority support.
Target Markets
- Enterprise AI Teams
- Financial Services
- Automotive
- Healthcare
- Government
- LLM Security & Safety
- EU AI Act Compliance
- AI Agent Performance Monitoring
- Computer Vision Model Validation
- L'Oréal
- AXA
- BNP Paribas
- Michelin