Privacy Policy

This policy explains what NextJet (operated by The New World Agency) collects, what it does not, and why. We have tried to write it in plain English. If anything is unclear, email jayden@thenewworldagency.com and we will explain it.

The short version

The automation runs on your own computer. Your Nextdoor login, your browser profile, the replies you send, and your voice file all stay on your computer, they are never uploaded to us. What we receive is the information needed to run your account, bill you, keep the product working, and power your dashboard: counts, timing, error messages, a copy of the leads the automation has already delivered to you, and, so you can review and correct the AI's decisions, a short snippet of the other posts the automation read, without the neighbor's name or neighborhood. Those review snippets auto-delete after 90 days and are used only to improve your own automation. By default, the AI that reads posts and drafts replies also runs through our own relay — see “How the AI reads your posts” below for exactly what that sends us and what we do (and don't) keep from it.

What we collect

• Account information. The email address and password you create at signup, and your business name. Passwords are stored scrambled (a “hash”) so even we can't read them, never in plain text.
• Business profile. The details you enter in the setup form (services, voice samples, guardrails, brand marks). This is used to build your automation bundle and is editable from your dashboard.
• Billing information. Handled by Stripe. We store a Stripe customer ID and your subscription status. We never see or store your full card number.
• Telemetry. The automation reports counts and timing (for example: number of scans, number of leads delivered, query performance) and error messages so we can spot problems.
• Delivered leads. When the automation alerts you about a lead, it sends a small amount to our servers to power the leads section of your dashboard, the public Nextdoor post link, the draft replies it wrote for you, and the full text of the post so you can read exactly what the neighbor asked. It does not pull out the neighbor's name or neighborhood as separate data, though the post is the neighbor's own public words and can include a name they wrote into it. See “The one exception” below for exactly how this works.
• AI-decision review snippets. So the automation's judgment can be checked and kept sharp, it also sends a short snippet(capped at 200 characters) of posts it read but did not deliver, along with the AI's verdict and reason and the post's public Nextdoor link (the same link anyone on Nextdoor can open). These snippets are included in your account data export. It never sends the neighbor's name or neighborhood, and never the full post. These snippets auto-delete after 90 days and are used only to improve your automation.
• Teaching messages. When you teach the automation, from your dashboard or by replying to a lead alert on Telegram, the message you write is sent to us and stored to personalize your replies. Neighbor post content is never included, and you can ask us to remove these at any time.
• Anti-abuse & security signals. At signup we record your IP address, a browser fingerprint, and normalized business name/website to prevent duplicate-trial abuse. For security and license enforcement we also log the IP address and a device identifier when your automation verifies its license, and when you download your bundle or attempt to log in. These are retained only as long as needed for fraud-prevention and security.
• Website usage (analytics). When you browse our public website, we collect anonymous, cookieless usage data, which pages you view, how long you stay, how far you scroll, what you click, and the site that referred you, tied to a random identifier kept in your own browser, never to your name or email. See “Website analytics” below.

What we never collect

• Your Nextdoor username or password. The automation uses the login you set up on your own machine, and it never leaves it.
• The stored contents of any AI relay prompt. By default, a classification/reply-drafting prompt (which contains the post text and can include whatever name or neighborhood the neighbor wrote into their own post) passes through our servers on its way to the AI provider, but we do not store or log it, no database record of it is ever kept. See “How the AI reads your posts” below. (If you switch your automation to subscription mode instead, that prompt never reaches our servers at all.)
• The neighbor's name or neighborhood as separate, saved data. The AI-review snippets described above are capped at a short excerpt and never include either, and they auto-delete after 90 days.
• The replies you actually send, unless you choose to paste one into the teach box to train your voice (see “Teaching messages” above). We also never receive the voice file (BRAND.md) on your machine, it is generated from the answers you gave us in the setup form, but the file itself is never uploaded back to us.
• Your browser profile or local lead database.

Lead content that reaches our servers

We want to be precise about this, because it is where anything lead-related reaches our servers. The automation scans on your machine and decides locally which posts are real leads. The winners, typically a handful per day, are delivered to you on Telegram (or Slack or email, if that's the channel you connected), and for those we mirror a record to our servers so your dashboard can show them: the public Nextdoor post link, the draft replies we wrote for you, and the full text of the post so your dashboard shows exactly what the neighbor asked. For the posts it read but did not deliver, we keep only a short snippet (capped at 200 characters) plus the AI's verdict, so you can review and correct its decisions, and those auto-delete after 90 days.

We deliberately do not extract or store the neighbor's name or neighborhood as separate fields. Because a delivered post is the neighbor's own public words, it can include a name or detail they wrote into the post themselves; we don't parse it out, save it as a contact, or use it to identify anyone. The rest of that information is another person's, and it stays on your machine and on Nextdoor itself (the post link opens the real post). What we do keep, we use solely to display your leads to you, to support your account, and to keep the Service reliable, we do not sell it, share it with other customers, use it for advertising, or build any product from it. Our endpoint enforces hard size and rate caps so this channel can only ever carry the few leads actually delivered to you. When your account is deleted, these records are deleted with it.

This is separate from the AI relay described next: this section covers the small, permanent mirror record your dashboard displays; the relay below covers the classification/reply-drafting prompt itself, which passes through our servers but is never stored.

How the AI reads your posts

By default, every automation reads posts and drafts replies through NextJet's own AI relay. The prompt, which contains the post text and can include whatever name or neighborhood detail the neighbor wrote into their own post, is sent from your computer to our servers and relayed to an AI provider on our own account: Anthropic, and/or OpenRouter, Inc. (a model-routing service we use so we can tune which underlying model handles a given request, fleet-wide, without requiring an engine update).

We do not store or log the contents of a relayed prompt or its response, there is no database record of what was inside it. What we do keep, so we can meter usage and keep it within a fair daily cap, is a count of calls and an estimated cost per license, nothing about what any individual prompt said.

One reliability footnote: if our relay is briefly unreachable and you happen to have the Claude command-line tool installed and signed in on your machine, the automation falls back to your own Claude subscription for that call, the prompt then goes straight from your computer to Anthropic and skips our servers entirely. If you don't have Claude installed (most customers don't), the automation simply retries on its next scan.

If you'd rather the AI run entirely on infrastructure you control, you can switch your automation to subscription mode, an environment setting that points it at your own Claude subscription instead. In that mode the prompt goes straight from your computer to Anthropic, exactly as if you pasted it into Claude yourself, and neither the prompt nor the post ever touches our servers. Subscription mode is optional and is not the default; switching to it is a change you make in your automation's settings, not something we do for you.

Cookies

We use essential cookies to keep you logged in to your dashboard (HttpOnly and secured), plus one long-lived first-party cookie set at signup that holds a random device identifier used purely to detect duplicate-trial abuse, never for advertising. We do not use third-party advertising cookies, social media pixels, or third-party analytics trackers on the platform. For analytics we use our own first-party, cookieless approach with a small amount of browser local storage, described in “Website analytics” below.

Website analytics

To understand how our public website performs and where visitors get stuck, we run our own first-party analytics, built and hosted by us, with no third-party trackers, no advertising networks, and no cross-site tracking. It works without cookies: instead, a random, anonymous identifier is stored in your browser's local storage so we can tell a returning visit from a new one. Unless and until you create an account, that identifier is not linked to your name, email, or any account (see the note below).

For each visit we record things like the pages you view, the page that referred you, how long you spend, how far you scroll, and which links or buttons you click (we strip anything that looks like an email address out of this). We do not store your IP address with this analytics data, an IP is used only momentarily, in memory, to rate-limit abusive traffic, and is not written to the analytics record. We use this information only in aggregate to improve the website; we never sell it or use it for advertising.

Because the identifier lives in your own browser, you stay in control: clearing this site's data (or visiting in a private window) removes it, and a new one is only created if you come back. We keep this usage data only as long as it is useful for understanding and improving the Service, and we may periodically delete older records.

One exception to “not linked to an account”: if you go on to sign up, we associate that first-touch information, the page and any campaign link that brought you, and the anonymous visitor identifier, with your new account, so we can understand which marketing actually leads to signups. We use it only for that internal attribution; we never sell it or use it to advertise to you.

Third parties we use

We rely on a small number of trusted providers to run the service: Stripe (billing), Anthropic (by default their models power your automation's AI through our own relay, on our account, described in “How the AI reads your posts” above; Anthropic also processes a few server-side requests for us directly, such as reading your business website to pre-fill setup and answering support-chat questions; if you switch your automation to subscription mode, Anthropic instead powers it on your own machine via your own separate Claude subscription), OpenRouter, Inc. (what: an AI model-routing service; why: when your automation uses our default relay, we may route a classification or reply-drafting prompt through OpenRouter to reach the underlying model instead of calling Anthropic directly, depending on how we're tuning model quality fleet-wide at the time; OpenRouter sees the same prompt content Anthropic would in that case, and is bound to the same no-training, no-retention handling we require of any AI provider we use), Resend (transactional email), Telegram (which delivers your lead alerts, and which we use to notify ourselves of key account events), Slack (only if you connect a Slack webhook, in which case we post the same lead cards — post link, your drafts, and the post text — to the workspace you chose), and our hosting and database providers Vercel and Supabase. Each processes only the data needed for its function, and none of them are permitted to use your data for their own purposes.

How we use your data

• To create and run your account and build your automation bundle.
• To bill your subscription and manage your trial.
• To show you your delivered leads and account metrics in your dashboard.
• To provide support and keep the product reliable.
• To prevent fraud and abuse of the free trial.
• To comply with legal obligations.

We do not sell your personal information, and we do not “share” it for cross-context behavioral advertising as those terms are defined under the California Consumer Privacy Act. We do not use lead content for advertising or for training any model.

When we may disclose data

We disclose data only to the service providers above, or if we are required to by law (for example, a valid subpoena or court order), to enforce our Terms of Service, to protect the rights, safety, or property of our customers or the public, or as part of a merger or acquisition of our business, in which case this policy continues to apply to your data and we will notify you of any change in ownership.

Data retention

We keep account and billing records for as long as your account is active and as required for legal and accounting purposes. Delivered leads are retained while your account is active so your dashboard history works. If you cancel and ask us to delete your account, email jayden@thenewworldagency.com and we will remove your personal data and your delivered leads, except where we are required to retain records. The lead and post history on your own computer is yours to keep or delete.

Security

Passwords are hashed, billing runs through Stripe, all traffic is encrypted in transit, and our servers only accept the exact short fields listed above and automatically reject anything more. No system is perfect, but the architecture is designed so that your most sensitive data, your Nextdoor login and your browser profile, always stays on your machine. By default, the full text of the posts the automation reads passes through our AI relay to reach the AI provider, never stored or logged on our end, see “How the AI reads your posts” above, unless you've switched your automation to subscription mode, in which case that text stays on your machine and goes straight to Anthropic through your own subscription instead. For delivered leads, the things that reach our servers are the public post link, your own draft replies, and the full text of the post; for the posts it didn't deliver we keep only a short excerpt (capped at 200 characters) plus the AI's verdict so you can review its calls, auto-deleted after 90 days. In neither case do we pull out the neighbor's name or neighborhood as separate data, though a delivered post, being their own public words, may include a name they wrote.

If a data breach ever affects your personal information, we will notify you without undue delay after becoming aware of it, and will notify regulators where the law requires it. We'll tell you what happened, what data was involved, and what steps you can take, by email to the address on your account.

Your rights

You can access, correct, export, or delete your personal data by emailing jayden@thenewworldagency.com. We will respond within the time required by applicable law. Depending on where you live, you may have additional rights under laws such as the GDPR or CCPA, including the right not to be discriminated against for exercising them. We honor those requests.

Children

The Service is a business tool for adults. It is not directed to anyone under 18, and we do not knowingly collect personal information from children. If you believe a child has provided us personal information, email us and we will delete it.

International users

We are based in the United States and process data on servers in the United States. If you access the Service from outside the U.S., you understand that your information will be transferred to and processed in the U.S.

Changes

If we update this policy we will change the date at the top. For material changes we will notify you by email before they take effect.

Contact

Questions about privacy? Email jayden@thenewworldagency.com.