Introduction
Giskard is headquartered in Paris, France founded in 2021 and is covered in Brel’s AI in Governance sector. Leadership referenced in company materials includes Alex Kombis.
Public company materials describe the firm as building AI software for governance workflows. Brel’s working summary of the product scope: Provides open-source and enterprise quality assurance testing environments for ML models.
This profile uses only company-linked sources captured during the 2026-07-10 import review. Funding, valuation, headcount, and ranking claims from third-party listicles are omitted unless a company-specific URL in the source list supports them.
What the company does
Giskard positions its product around the following use case: Scanning LLMs and predictive models for bias data leakage and hallucinations. Secondary focus noted in research intake: Model Testing.
Buyers should treat vendor marketing claims as unverified until confirmed in a live evaluation against their own data, compliance, and integration requirements.
Who it serves
Stated buyers and users include: AI engineers and quality assurance teams
Procurement teams should confirm whether the deployment model (SaaS, on-prem, hybrid) and data-residency options match their sector obligations before pilot expansion.
Company background
Giskard operates from Paris, France. Company materials cite 2021 as the founding year. Alex Kombis is listed as founder or CEO in the intake sources used for this profile.
Brel does not infer prior employers, investor syndicates, or competitive rankings without a company-specific citation.
Product and AI capabilities
Within Governance AI, Giskard concentrates on: Scanning LLMs and predictive models for bias data leakage and hallucinations.
Exact model architecture, training data rights, and evaluation harnesses are rarely fully disclosed on marketing pages — request technical documentation during vendor diligence.
Key developments
Recent development noted in intake (verify on the linked company page before citing externally): Deployed interactive testing loops for collaborative model vulnerability diagnostics.
Why it matters
Giskard matters for Brel readers tracking Governance AI because it represents an active vendor in this category with a verifiable public website and company-linked documentation. Its stated focus — Scanning LLMs and predictive models for bias data leakage and hallucinations. — is a concrete buying motion teams can compare against peers in the same sector hub.
Use the AI in Governance hub to compare adjacent profiles; do not treat directory presence as an endorsement.
Sector context
Giskard is filed under Brel’s AI in Governance coverage, with secondary theme Model Testing. Country taxonomy reflects headquarters (France), not customer geography.
Sources and references
Primary references used for this profile:
- Giskard — official source — https://www.giskard.ai/product
- Giskard — supporting company page — https://www.giskard.ai/
Last reviewed 2026-07-10. Request a correction via the form linked on this page if a fact is outdated.