Privacy Policy

Effective date: October 20, 2026 Replaces: the policy dated October 8, 2020

Who we are

This site is operated by Hyper Vigilance, LLC, a Virginia limited liability company — the same statement of identity the site's footer carries. Questions about this policy go to the team mailbox: engage@hypervigilance.com.

The short version

This site collects personal information in exactly one place: the form on /engage/. It sets no cookies of its own, runs no advertising or remarketing, sells nothing about you, and hands your details to no outside marketing or sales platform. What you submit goes to a queue on our own hosting, then into our own records, and is deleted on the schedule published below. If anything you write into the form looks like controlled information, we do not keep that field or your note, and the page tells you exactly what was not kept.

The rest of this document is the detail behind those sentences.

What we collect, and where

The form on /engage/ is the only place this site asks you for anything. It collects:

• your name, work email, and company;
• what prompted your search and how you found us (each picked from a fixed list — you cannot type free text into either);
• anything you choose to write in the optional note field (limited to 300 characters);
• only if you fill them in, three optional answers about when a call would suit you: a time-of-day band, a US time zone, and a rough time horizon — each picked from a fixed list, with no free text.

No phone number is asked for and none is required. There are no accounts, no registration, no login, and no other form anywhere on the site.

The form also carries three technical signals that are not about you as a person: a hidden trap field that only automated form-fillers complete; when your browser runs our page script, a count of how long the page was open before you pressed submit; and a flag recording that the script ran at all. These exist to score automated spam. None of them prevents a real person's submission from going through, and if your browser does not run scripts, the absence of these signals is not held against you.

What a submission record contains besides your answers. When you submit, the stored record also carries: the time of submission; a random reference identifier we generate for that submission (it identifies the submission, not you, and connects to nothing else); the version of the form you used; the address of the page you submitted from, and the address of the page that referred you here if your browser supplied one — each reduced to its site and path, never the query string that may have followed it; and the spam-scoring outputs (the bot-check verdict when the bot check is enabled, a trust score, a review flag, and which of the technical signals fired). We use these to operate the intake and review suspicious submissions, not to profile you.

Why those two addresses are cut, and where the cut happens. A link decides what a query string carries, not the person following it: an address of the shape /engage/?ref=partner&token=... can hand us a session token or a third party's email address without you ever typing one. So both addresses are reduced to site and path at intake — on our side, inside the code that assembles the stored record — before anything is written down. What is kept is the site and the path; the query string and any fragment are discarded and never stored. Because the cut lives in the code that builds the record rather than in the page script, it does not depend on your browser: a submission sent with scripting switched off is cut exactly the same way. If an address reaches us in a form we cannot read, we store nothing in its place rather than storing the unreadable text.

What we refuse to accept

This form is a commercial, unclassified channel, and the page says so before you type. The note is checked for the shape of controlled information: CUI or FCI markings, classification banners, export-control references, distribution statements, contract or order identifiers, registered entity identifiers. Your name, work email, and company are checked for markings and identifiers only, so that an ordinary name or address, such as the surname Cui, is not mistaken for one. A field that matches is not kept: its text is not written to storage, not logged, and not echoed back. When any field matches, the note is not kept either, even if the note itself was clean. The two page addresses described above are checked like the note, and one that matches is not kept. What we keep is every contact field that did not match, your answers from the fixed lists, and the category of what was seen (for example "a contract identifier") — never the text itself. If the field that matched is your work email, we have no address to reply to, so the page gives you the team mailbox instead. The refusal message on the page names exactly which fields were not kept and what to do next.

Be clear about what this check can and cannot do. It matches the shapes of markings and identifiers. Controlled detail written as plain prose, with no marking or identifier shape, cannot be detected by any automated check — it would be stored as an ordinary note under the 90-day rule below. That is why the instruction on the page is not "we will catch it" but "do not put controlled detail in this form at all." If you need to convey controlled information, tell us only that a conversation is needed; we will set up a channel that can take the detail properly.

What we measure about reading, and what we deliberately do not

Pages on this site measure, in your browser, how long each section of a page was actually being read — the clock runs only while the section is on screen, the tab is visible, the window is focused, and you are not idle. What is recorded per section: engaged milliseconds, how far into the section you scrolled, and how many times it came into view. Each batch also carries the page's path on this site, a random per-view identifier, the session identifier described below, whether your system asks for reduced motion (so we can judge reading behavior without animation bias), and the referral class described next.

What this measurement excludes, by design and enforced at the data contract: your IP address, user agent, screen or viewport size, timezone, language, query strings, and the address of the page that referred you — the measurement keeps only whether you arrived from inside this site, outside it, or directly. (That three-value reduction is the reading measurement's own. The intake form on /engage/ keeps a little more, as disclosed above: the site and path of two addresses — and no query string there either.) The identifier attached to a batch is a random value held in your browser's session storage — it is not a cookie, it is not persisted, it is cleared when the tab closes, and it cannot connect your visits across sessions or across sites.

If your browser sends the Do Not Track signal or the Global Privacy Control signal, this measurement does not run at all — nothing is measured and nothing is sent.

These batches go only to this site's own address, never to a third party. As of this policy's effective date there is nothing behind that address: no collector receives the batches and none of them is stored anywhere. Two commitments follow. If collection is ever switched on, this section gains a stated retention period before that happens. And the measurement's data contract permits optional content-position markers that are never emitted — the build strips them and a gate proves it — so if those are ever enabled, this section is amended first.

Hosting, and what the infrastructure sees

The site is served by Cloudflare Pages; Cloudflare is our hosting provider. Like any host, Cloudflare processes your IP address and request details to deliver pages and protect the service, and Cloudflare's own privacy policy governs what it does with them.

Our own intake endpoint uses your connecting IP address for exactly one thing: a rate limit that stops a script from flooding the form. The address itself is never stored — it is reduced to a salted, truncated one-way hash, and the counter it feeds expires within one hour. If the rate limiter's storage is unavailable, submissions proceed rather than being dropped.

We do not log the content of what you type in server logs.

The bot check (only if enabled)

The form may load a bot check from Cloudflare (Turnstile). As of this policy's effective date it is switched off and the page loads nothing from outside this site. If it is enabled: your browser talks to challenges.cloudflare.com, and our server verifies the result with Cloudflare. The check is advisory only — if your network blocks it, or it fails, your submission still goes through; a failed or missing check affects only an internal review score. This is stated on the form itself.

Analytics (only if enabled)

As of this policy's effective date, this site loads no outside analytics — no Google Analytics, no pixels, no session recording.

The build carries a switch that can enable Google Analytics 4. It is off unless deliberately enabled by the operator. If it is ever switched on, this section is updated to say so before that change ships: GA4 would load from www.googletagmanager.com, set its own cookies, and process usage data under Google's privacy policy, with IP anonymization requested.

Cookies

This site sets no cookies of its own. There is nothing to consent to and no cookie banner because there are no cookies to disclose — the reading-time measurement above uses session storage, which is not a cookie and does not persist.

If the bot check or analytics described above are enabled, their providers (Cloudflare, Google) may set cookies of their own; that is the only way a cookie appears on this site, and this section must be updated before either is switched on.

Where your information goes

A submission travels one path: a durable queue on our own hosting, then our own records system, then a notification to the team mailbox (engage@hypervigilance.com). Replies to you come from that shared mailbox.

• It is not handed to an outside marketing or sales platform.
• It is not used for advertising, and we run none.
• It is not sold, rented, or shared for anyone else's marketing.
• There is no third-party scheduling vendor: if you tell us when you are reachable, a person replies from the team mailbox with specific times.

We use your submission to reply to you and to conduct the conversation you asked for. We do not add you to a newsletter or marketing list.

How long we keep it

These periods are published on the form page itself and are a commitment, not a preference:

Content refused by the controlled-information check is not retained for any period: the field it was typed into is stored empty or as a fixed placeholder, and the refused text is not written to the queue, our records, our logs, or the notification to the team mailbox. Backup copies of the queue and of our records roll off within 30 days, so anything deleted on the schedule above is gone from those backups within 30 days of its deletion.

Email

Mail to and from engage@hypervigilance.com is ordinary business correspondence, held in our own tenant and covered by the business-records line above once a conversation becomes an engagement. We send you email only to conduct the conversation you started — never marketing broadcasts.

Your choices and rights

• Write to engage@hypervigilance.com to ask what we hold about you, to correct it, or to have it deleted. Absent a legal reason to keep something, we will honor a deletion request ahead of the schedule above.
• The reading-time measurement honors your browser's Do Not Track and Global Privacy Control signals automatically — no request needed.

Security

We protect this data with the same discipline we sell, and we will not pretend otherwise is impossible: no method of transmission or storage is completely secure. What we can honestly say is that we minimize what exists to be lost — the shortest path, the fewest copies, published deletion schedules, a refusal to store the most dangerous category of content at all, and addresses cut to site and path before they are written down.

If you believe you have found a security flaw in this website, our vulnerability disclosure policy tells you how to report it and what we commit to in return.

Disclosures we may be required to make

We may disclose personal information if required by law or valid legal process, or to protect against legal liability. If the business is merged or acquired, submissions and records may transfer with it; this policy's commitments would follow the data.

Other websites

Where this site links outward — standards bodies, official sources — those sites have their own privacy practices, and this policy does not cover them.

Children

This site is a business-to-business service directed at companies in the US defense industrial base. It is not directed at children, and we do not knowingly collect information from anyone under 13.

Changes to this policy

If this policy changes, the change is published here with a new effective date. The retention periods above are also printed on the form page itself; changing a period means changing that page too.

Contact

engage@hypervigilance.com