AI Governance & Security Vendor Wiki

6 min read

Holistic AI

One-liner — An AI governance, risk and compliance (AI GRC) platform that inventories an organization’s AI systems, runs technical risk/bias tests, and maps them to regulations (EU AI Act, NIST AI RMF, ISO 42001, NYC Local Law 144); known originally for independent bias audits of hiring algorithms.

Categories — ai-governance-platform

What it does

Holistic AI is an end-to-end AI governance platform. It does three jobs: (1) discover and inventory AI systems across cloud, code repos and SaaS (to surface “shadow AI”); (2) assess risk with a library of technical tests for bias, robustness/safety, security and performance; and (3) map and enforce compliance against regulatory frameworks with control-mapping, gap analysis, reporting and approval workflows. The company started in the algorithmic audit niche — it is one of the recognized firms (alongside BABL AI, Conductor AI) that performs independent bias audits of Automated Employment Decision Tools (AEDTs) to satisfy NYC Local Law 144, producing the public “Summary of Results” and fairness impact ratios that law requires. It has since broadened into a full governance/registry platform.

Where it sits in the stack

This is a ai-governance-platform play at the governance layer — the system-of-record and policy/oversight tier that sits above the runtime data path, not in it. Its job is inventory, documentation, risk assessment, attestation and regulatory mapping, not inline traffic inspection.

  • Runtime-control role: none directly. Holistic AI is not an inline prompt/egress control — it is an oversight/registry layer. (Its newer “Guardian Agents” / AI Safeguard runtime features edge toward monitoring, but the core product is governance, not a firewall.)
  • Trust zone: spans the governance plane across all zones; it documents and risk-rates systems wherever they run rather than enforcing a boundary between red/yellow/green.

Deployment & architecture

SaaS platform with connectors/integrations into cloud and dev tooling (AWS, Azure, GitHub, Databricks; “20+ integrations” per vendor) for automated discovery, plus an API. Bias-audit engagements are delivered as a service with results hosted on a dashboard and a public summary page. Deployment specifics (e.g., self-hosted/on-prem options) are not clearly documented publicly — treat as SaaS-first.

Positioning & differentiators

Holistic AI’s distinguishing angle is regulatory/audit depth: it grew out of academic algorithm-auditing work at UCL and leans on standards credibility (its solutions appear in the OECD Trustworthy AI Toolkit; it claims Observer Status with the Council of Europe Committee on AI — vendor claim). Versus its neighbors:

  • credo-ai — closest competitor; both are governance/registry + framework-mapping platforms. Credo AI is more policy/governance-workflow led; Holistic AI carries more technical-testing and formal-audit heritage.
  • modelop — more focused on operational model governance / ModelOps for production model fleets (closer to SR 11-7 model inventory and monitoring in banks).
  • ibm-watsonx-governance — incumbent/stack play bundled with IBM’s AI platform; heavier, enterprise-IT anchored.
  • monitaur — model-risk/assurance focused, strong in insurance/regulated ML.
  • fairly-ai — overlapping AI GRC / compliance-automation focus, smaller.

Holistic AI’s bias-audit lineage makes it relatively strong on fairness/EEO use cases (hiring) where others are thinner.

Ownership, funding & M&A

  • Ownership: independent, private, VC-backed. No acquisition of the company was found. Confidence: medium (multiple sources agree it remains independent; Tracxn states no M&A).
  • Founded: 2020, out of University College London, by Emre Kazim and Adriano Soares Koshiyama (co-CEOs). HQ London, with US presence in Palo Alto.
  • Investors: Tola Capital, Mozilla Ventures (investment announced 2024-03-18), Premji Invest, Innovate UK, Microsoft for Startups Pegasus Program, Kickstart Accelerator.
  • Funding total: uncertain. A figure of ~$35M (venture, ~May 2024) is repeated by data aggregators, but that date collides with unrelated news and is low confidence. Earlier Innovate UK grant of ~£596,670 for an AI risk-management platform is documented.

Name-collision caution (not a contradiction): A separate, unrelated Paris-based AGI startup originally called “Holistic”/“Holistic AI” rebranded to “H” and raised ~$220M seed (DeepMind alumni; Accel) in May 2024. Crunchbase/PitchBook/press results frequently conflate the two. The $200M+ figures and the DeepMind founders belong to H, NOT to this London AI-governance vendor. Do not merge these companies.

CTO / hedge-fund lens

  • Day-2, low fit for most funds. This is an enterprise AI-GRC platform whose sweet spot is organizations deploying many AI systems and facing direct AI regulation — large enterprises, HR/hiring-tool vendors, and EU-exposed firms. A 50-person fund does not need a discovery-and- attestation platform of this scope on Day 1.
  • SR 11-7 / model risk: adjacent but not a native fit. SR 11-7 is about model validation, documentation and independent review of quantitative/trading/credit models; Holistic AI’s governance registry and risk-assessment workflow can support that documentation discipline, but it is built around AI/ML system governance and regulatory mapping (EU AI Act, NIST AI RMF, ISO 42001), not SR 11-7 model-validation methodology specifically. Funds that already run a model-risk function (modelop, monitaur are closer matches) will find more overlap there.
  • When you’d actually need it: if the fund builds or sells AI-driven HR/screening tools (NYC LL144 exposure), has material EU AI Act obligations, or wants a single inventory + attestation system across a sprawling internal AI footprint. Otherwise it’s premature.

Competitors / alternatives

credo-ai · ibm-watsonx-governance · modelop · monitaur · fairly-ai

Open questions / to verify

  • Exact funding total and latest round — needs a primary press release; aggregator figures (~$35M) are conflated with the Paris “H” company. Currently low confidence.
  • Whether self-hosted / on-prem deployment is offered (only SaaS confirmed).
  • Current employee count / scale and revenue traction (unknown).
  • Substantiate the Council of Europe “Observer Status” claim independently (vendor-stated).

Sources

History

  • [2026-06-28] Stub created from seed registry.
  • [2026-06-28] Researched; established 2020 founding (UCL), London HQ, co-CEOs Kazim & Koshiyama, independent VC-backed ownership (Tola Capital, Mozilla Ventures Mar-2024, Premji Invest); confirmed AI-GRC/bias-audit positioning and EU AI Act/NIST/ISO 42001/NYC LL144 mapping. Flagged name collision with unrelated Paris “H” ($220M) startup — funding total left low confidence. hedge_fund_fit set low (Day-2).

Graph View

Backlinks