A governance firm should publish its own governance.

When you complete the Architecture Snapshot, the AI Readiness Assessment, or the Diagnostic intake, we collect what you provide: your name, business name, email, role, sector, headcount range, optional revenue range, and your responses. During a paid engagement, we also hold the documents and data you choose to share with us. That is the full list. We do not run advertising trackers, third-party analytics scripts, or marketing pixels on this site — none.

Your responses are transmitted over encrypted connections and stored in an access-controlled database encrypted at rest. Engagement files are kept on encrypted company devices and accounts. Access is restricted to Omnira and to the service providers below, which host, store, process and send it on our behalf, under data-protection terms:

• Netlify — hosts this website and receives your form submissions
• Supabase — the database where your responses are stored
• Resend — emails us a notice when you submit an assessment
• Google Workspace — our business email and calendar, and the spreadsheet where we keep track of our conversations with prospective clients
• Anthropic — Claude, for AI-assisted analysis and drafting, with model training switched off

Between ourselves, we message on Signal, which is end-to-end encrypted: not even Signal can read what we send. If our providers change, this list and the version number at the top of this page change with them.

Anything we email you, including a copy of your results, comes from our own business email. We do not use your address for anything you didn't ask for.

Your data is never sold, rented, traded, or shared with third parties for their own use. No "trusted partners," no fine print reversing this sentence. The only thing that overrides it is the law: if a court order or subpoena compels disclosure, we disclose only what is required, and we tell you first unless the law forbids it.

Email Sales@OmniraConsulting.com at any time to request a copy of what we hold about you or to have it deleted. We honor both promptly. The one exception is the record of a paid engagement — the signed agreement, your intake, invoices, the delivered report and engagement correspondence — which we keep for four years after delivery for legal, tax and contract reasons, and then delete.

If we learn that information you gave us has been accessed without authorization, we will tell you promptly, and always within the time the law requires.

Your assessment responses are used to score your instrument, generate your report, and inform any engagement you pursue with us. We may follow up with you about your results. That is the primary and dominant use of your data.

Anonymized responses may inform aggregate benchmarks — for example, how businesses in your sector and size band typically score. Before any response enters an aggregate, identifying information is removed. We never display a benchmark cell containing fewer than five businesses, so no individual company can be inferred from a comparison. If the data is too thin to say something honestly, we say nothing.

When a Diagnostic includes a confidential survey of a client's team, anonymity is achieved by architecture, not promise: we do not collect names, individual tracking tokens, or network identifiers; timestamps are rounded to the day; results are never broken out for groups smaller than five people; and open-text comments are synthesized in findings, never quoted in a way that could identify the writer. Leadership receives patterns, not people. We never give a client respondent-level data, and our engagement agreement forbids clients from trying to identify a respondent or acting against anyone for taking part.

Omnira uses AI tools in research, analysis, drafting, and development. This makes the work faster and, applied correctly, better. We say so plainly because it is true and because pretending otherwise would be a strange position for a firm whose founder designed one of the first enterprise generative AI governance programs.

Every deliverable that carries the Omnira name is reviewed, validated, and owned by the practitioner. AI accelerates the work; it does not replace the judgment you are paying for. Nothing is delivered to a client that has not been personally examined.

Client-identifying information is only processed in AI tools with model training switched off, and we never send feedback ratings on conversations that contain it. Today that tool is Claude, from Anthropic. Client confidentiality survives contact with our toolchain.

Any client may ask how AI was used in their engagement and receive a direct answer. The same standard we build into governance frameworks for clients applies to us first.

Omnira accepts no referral fees, commissions, reseller margins, or affiliate payments from any software vendor, platform, or service provider we recommend. When we name a tool in your engagement, the recommendation is based on your architecture — not our economics. The same holds for referral relationships: if we recommend a firm that also refers clients to us, we tell you before we make the recommendation. If that ever changes for a specific relationship, it will be disclosed to you in writing before any recommendation is made.

We do not concurrently serve direct competitors on overlapping strategic matters without the informed consent of both parties. Where a potential conflict exists, we disclose it and, if it cannot be resolved, we decline the work. Declining an engagement is cheaper than compromising one.

Any material financial interest Omnira or its principal holds in a company relevant to a client's engagement is disclosed before that relevance can influence a recommendation.

We report what we find, not what is easiest to hear. If a diagnostic reveals that your architecture is sound and your problem lies elsewhere, that is the finding you receive — along with our recommendation for who can actually help.

We never manufacture precision. Ordinal data is presented as ordinal, benchmarks are only shown where the sample supports them, and no chart in an Omnira deliverable implies certainty the underlying data cannot carry. A confident-looking number built on thin data is not analysis; it is theater, and we do not perform it.

What we learn inside your business stays inside your engagement, subject only to the legal exception in Policy I. Client identities and engagement details are not used for marketing without express written permission.

Deliverables belong to the client once paid in full — systems, documentation, and the understanding to run them. Omnira keeps its own methods, instruments and benchmarks; the client may use them as they appear in its deliverables, inside its own business. We do not build dependency into architecture and call it partnership.

Workforce optimization affects human beings. We present its realities without euphemism and without cruelty, we plan transitions responsibly, and the decision about freed capacity always remains the owner's. We will not dress a workforce decision in technical language to help anyone avoid owning it. Employment decisions should be reviewed by the client's employment counsel; Omnira does not give legal advice.

Any policy Omnira drafts for a client is designed to be reviewed and approved by the client's legal counsel before taking effect. We architect; your attorney ratifies. That sequence is non-negotiable, including when it slows things down.

Every engagement is fixed-scope and fixed-price, agreed in writing before work starts. There is no hourly billing.

The Diagnostic fee is paid in advance and is non-refundable once work begins. We are paid for the work, done to the scope in your agreement — not for agreement with what it finds. If we fail to deliver on time for reasons within our control, or end the engagement ourselves, we refund the fee in full. Your engagement agreement sets out the details.

The $7,500 Diagnostic fee is credited in full toward any Build or the AI Governance Program signed within 90 days of your report's delivery.