# ZeroH > Your Sovereign, Governed AI Workforce. AI employees, not AI freelancers: every job runs masked on your soil, > waits for your approval where it matters, and leaves proof an auditor can check. ## What ZeroH is ZeroH is one on-soil substrate that RUNS an organisation's AI, TELLS it how that AI is doing, and PROVES the AI is compliant. Four pillars: Build (compose the work), Run (the governed runtime), Know (intelligence and learning off the organisation's own evidence), and Prove (compliance as a byproduct rather than a project). It is built and operated by Blade Labs. ## The governed turn Every request through ZeroH takes the same path, and that path is the product. Personal data is masked on your own infrastructure BEFORE any model sees it, and an unknown category fails closed rather than passing through. Consequential actions wait for a human approval instead of being taken on the model's word. Every turn appends to an append-only, hash-chained ledger, and an action that cannot be recorded is refused rather than reported as done. The masked values are restored on your soil, and the restoration map is never written down. The trace this produces is the control, the intelligence and the evidence at once. ## Products - ZeroH platform: the governed AI workforce, departmental agents that hold jobs, a console for the people accountable for them, governed connectors into the systems the work already lives in, and proof packs a regulator can read. - Ask Ali: a Shariah-grounded AI copilot for Islamic finance, answering from a governed corpus with citations, on the same receipted spine as the rest of the workforce. ## Attestations - SOC 2 Type II - ISO 27001:2022 - GDPR - Fatwa Certified - UK Patent Pending ## Contact Book a 30-minute demo or talk to us at https://zeroh.io/contact. ## Full content The complete text of every page, in one file: https://zeroh.io/llms-full.txt. ## Pages - [ZeroH · Your Sovereign, Governed AI Workforce](https://zeroh.io/): AI employees, not AI freelancers. Every job runs masked on your soil, waits for your approval where it matters, and leaves proof an auditor can check. - [Strategic Intelligence, ZeroH](https://zeroh.io/intelligence): Adoption, quality, cost and learning read straight off the governed ledger: who uses the workforce, what it costs, what it learned, and every agent’s scorecard. - [Knowledge & Memory, ZeroH](https://zeroh.io/knowledge): How the workforce comes to know your work: meeting decisions captured the day they happen, a SharePoint folder turned into cited answers whose readers inherit the folder’s own permissions, and confidential shelves that stay dark until a person lifts them. - [Connectors & Tools, ZeroH](https://zeroh.io/connectors): What the governed workforce can reach, and how a new source becomes safe to say yes to: configure, probe, capability preview, approve, with the credential held in your key vault by reference and every approval on the ledger. - [Policy & Control, ZeroH](https://zeroh.io/policy): The rules that run before the model does: custom policies, spend caps enforced before the call, approval gates on consequential actions, per-field masking, model allowlists, every change versioned and audited, and a skill supply chain that never updates itself. - [Self-Improvement, ZeroH](https://zeroh.io/self-improvement): Observe, ratify, gate. Your workforce gets sharper from its own governed evidence, and the learning stays your IP, on your soil. - [Proof & Compliance, ZeroH](https://zeroh.io/proof-security): The receipt behind every AI action: framework readiness, the law traced to running code, and regulator-grade proof packs for Qatar/QCB, ISO 42001, GDPR and the EU AI Act. - [AI Budgeting, ZeroH](https://zeroh.io/ai-budgeting): Do not use a frontier model for a non-frontier problem: policy-routed models, open weights on your own GPU, evals that keep the routing cheap, and hard budget caps enforced before the model is called. - [Deployment, ZeroH](https://zeroh.io/deployment): SaaS-first, land and expand: start on the governed SaaS in a day, move to the managed app on your own Azure when you are ready. Same spine, same receipts. - [Pricing, ZeroH](https://zeroh.io/pricing): Free platform on our fleet: $0.00 a month, pay-as-you-go AI at the provider's price with 0% markup through your Microsoft bill. Your-own-tenant deployment priced per agreement, by contact. - [Ask Ali, Islamic Finance AI Copilot, ZeroH](https://zeroh.io/ask-ali): The Shariah-grounded AI copilot: cited answers from a governed corpus, on the same receipted spine as the rest of your workforce. - [Contact, ZeroH](https://zeroh.io/contact): Book the 30-minute demo, or talk to us about the governed AI workforce. - [By role, ZeroH](https://zeroh.io/for): What the governed AI workforce changes depending on who you are: the CISO signing off on delegated access, the DPO who needs masking before the model, the AI risk officer, the compliance officer, the Shariah board and the founder watching the bill. - [ZeroH for CISOs](https://zeroh.io/for/ciso): Least-privilege AI with an append-only ledger: what the CISO signs off on, and why. - [ZeroH for Data Protection Officers](https://zeroh.io/for/dpo): Host-side masking before any model, consent-scoped reads, and a console built for the DPO. - [ZeroH for AI Risk Officers](https://zeroh.io/for/ai-risk-officer): Model risk with receipts: policy-routed models, human gates, and the trace behind every run. - [ZeroH for Chief Compliance Officers](https://zeroh.io/for/compliance-officer): Compliance as a byproduct of running the work: proof packs, control traces, framework readiness. - [ZeroH for Shariah Board Members](https://zeroh.io/for/shariah-board): Cited, corpus-grounded answers and an auditable trail for every fatwa-relevant judgement. - [ZeroH for Fintech Founders](https://zeroh.io/for/founder): Ship AI your regulator will accept: the governed workforce that comes with its own evidence. - [Patents, ZeroH](https://zeroh.io/patents): Virtual patent marking: UK application GB2604344.8, Systems and Methods for Cryptographic Disclosure Provenance of Self-Governing Data, filed 27 February 2026 by Blade Labs Holdings Private Limited, with the products and features it covers, and the notice that a filing is not a grant. - [Privacy Policy, ZeroH](https://zeroh.io/privacy-policy): How ZeroH processes personal data. - [Terms of Use, ZeroH](https://zeroh.io/terms-of-use): The terms governing use of ZeroH. - [Cookie Policy, ZeroH](https://zeroh.io/cookie-policy): How ZeroH uses cookies, and mostly does not. - [Code of Conduct, ZeroH](https://zeroh.io/code-of-conduct): The standards we hold ourselves to.