Privacy policy

Rivyt is operated by Naveiyra Holdings, Inc. dba Rivyt, a Delaware corporation. “We”, “us” and “Rivyt” in this policy mean Naveiyra Holdings, Inc.

Rivyt answers questions from your company’s own documents. We use what you give us to do that and nothing else. We do not sell your information, we never use your company’s content or your account for advertising, and we do not use your content to train AI models.

1. Who this covers

Rivyt is a workplace tool for manufacturing and industrial companies in the United States. Your employer (“your company”) signs up, invites you, and decides what documents Rivyt can read. Rivyt is available only by invitation from a company. It is not a consumer product.

There are two roles in this policy, and they matter for your rights:

• For your company’s content (its documents, the questions people ask, the answers Rivyt gives, and the knowledge Rivyt builds from them), we act on your company’s instructions as its service provider. Your company decides what happens to that content. See section 9.
• For your account and for how you use the app, this policy explains what we do directly.

This policy covers the Rivyt app at app.getrivyt.com, Fox (Rivyt on a phone, by voice), and this website, getrivyt.com.

2. What we collect, where it comes from, and how long we keep it

This is our notice at collection. The categories on the left are the ones California law uses. We collect nothing outside this table.

We keep something longer only when the law requires it, for example to answer a legal claim, and only for as long as that lasts. Backup copies are deleted as those backups expire.

What we do not collect: your precise location, biometric identifiers (we make no voiceprints), government ID numbers, financial account numbers, health information, or information about your race, religion, sexual orientation or other protected characteristics. If your company’s documents happen to contain any of these, they are your company’s content, handled as in the table above.

What stays in your browser: in the Rivyt app, only what keeps you signed in, and your display preferences (theme and default screen). The app uses no advertising cookies, analytics trackers or tracking pixels. This website is covered in section 2A.

2A. This website (getrivyt.com)

Visiting getrivyt.com does not require an account. Our hosting provider keeps standard request logs, such as your IP address and browser. Beyond that, we collect only what you choose to send us:

• Request access. Your work email, company and anything you write. It is sent to us by email through Formspree, so we can reply.
• Guide downloads and the pilot waitlist (on our pages for people arriving from an ad, and from a LinkedIn lead form). Your name, work email, company, role, phone number if you give it, whether you asked about a pilot, and how you reached us (the ad campaign and click identifiers in the link, the page you landed on, the page that referred you, and your browser). We store it in our database (Supabase), email it to ourselves (Formspree), and add you to our sales contact list (Apollo), so we can send the guide and follow up once. We do not sell it or add you to a newsletter.
• Advertising measurement, on ad landing pages only. The pages under getrivyt.com/lp, which people reach from an ad, load the tag of the advertising platform that showed the ad (Google Ads, Meta or LinkedIn). The tag sets cookies and tells that platform that you visited and whether you asked for the guide, so we can tell which ads work. The tag is not loaded if your browser sends a Global Privacy Control signal. No other page on this website, and nothing in the Rivyt app, loads an advertising tag.

We keep sales enquiries while we are in touch with you, and delete them if you ask. Email privacy@getrivyt.com.

3. How we use it

We use personal information only to:

• answer your questions from your company’s documents and show the sources;
• suggest questions that fit your work, using your job title, your profile and the equipment you ticked (leave them blank, or switch tailoring off, and the suggestions are general instead);
• tell colleagues the identity of who knows about a piece of equipment, when you have indicated that you do;
• record the knowledge you share in an interview, attributed to your role, for your company’s admin(s) to review before it is used in answers;
• keep accounts secure, prevent abuse, and fix things that break;
• send sign in, invite and security emails;
• answer support requests, privacy requests and feedback;
• meet our legal obligations and enforce our agreements.

We do not use it for advertising, we do not build profiles of you for any purpose, and we do not use it to make decisions about you that have legal or similarly significant effects.

Who at your company can see what. People in your company can see your profile. Your company’s admins can see the question history of members of the workspaces they belong to. A company admin who joins a private workspace must give a reason, and that workspace’s members can see it.

Rivyt staff. Rivyt staff read a customer’s content only to give support the customer asked for, to keep the service secure, or to comply with the law, and only with a stated reason.

4. Who processes it

We use these companies (our service providers, sometimes called subprocessors) to run Rivyt. Each receives only what its job needs, each processes it only to provide its service to us, and none of them may use your content to train their models. Where each one stores and processes data is in section 13 and on our subprocessors page.

In the past 12 months we have disclosed the categories in section 2 to these service providers, for the purposes in section 3, and to no one else.

Changes to this list. When we plan to add or replace a company on this list, or change what one receives, we email every customer’s company admins before the change takes effect, and update our subprocessors page. Anyone else can subscribe to the same email.

Other disclosures. We disclose information if the law requires it (for example, a valid subpoena or court order), and we tell your company unless the law forbids it. If Rivyt were sold or merged, information would pass to the buyer, who would remain bound by this policy.

5. Google Drive and Google user data

This section explains what Rivyt does with data it receives from Google when a company connects Google Drive: what it accesses, how it uses, stores and shares it, how it protects it, and how long it keeps it. A summary is on our Google data use page.

What we access

A company admin can connect a Google Drive account to a workspace. When the admin connects with Google’s file picker, Rivyt asks Google only for:

• Access to the files you pick, and only those (Google’s drive.file permission). An admin chooses files in Google’s own file picker. Google then lets Rivyt read those files and notice when they change. Google does not let Rivyt see, list or search anything else in the Drive. The same permission covers files Rivyt itself creates; Rivyt cannot open, change or delete any other file.
• The email address of the Google account used to connect (Google’s sign in permissions), so admins can see which account is connected.

Connections made before the file picker. A Drive connection made before Rivyt moved to Google’s file picker asked Google for read access to the Drive and to Google Sheets, and for the same permission for files Rivyt creates, so that an admin could browse the Drive in Rivyt and choose folders or files to sync. With such a connection, Rivyt reads the contents only of the folders and files an admin chose, including files later added to a chosen folder. Each such connection is replaced when its workspace reconnects with the file picker, and its broader access is then revoked at Google.

For each file it reads, Rivyt also reads the file’s name, its place in Drive and, where Google provides them, its sharing settings, including the email addresses and groups it is shared with.

How we use it

Rivyt reads the files and turns them into text, splits the text into passages and indexes them, and builds its knowledge graph (“Rivyt Mind”) and its record of events from them. It uses all of this only to answer questions from people in that company’s workspace, with a citation back to the file. Answers drawn from a file are available to the members of the workspace it was added to, so an admin should add only files that everyone in that workspace may read. Rivyt keeps each file’s sharing settings with its record, and uses them only within the service.

How we store it

In Supabase (our database and file storage): a copy of each file, with every earlier version as it changes; the text extracted from it; the knowledge graph facts and events drawn from it, each with a quotation from the file as evidence; and its name, location and sharing settings. In Pinecone (our search index): the passages and their numeric representations, with the file’s name and heading, in a separate area for each workspace. Both store this data in the United States. The Google access tokens are encrypted at rest by Supabase, and encrypted again by Rivyt with a key held separately in Supabase’s key vault.

Who we share it with

Only the service providers in section 4, each only for its job:

None of them may use it to train their models. Anthropic, Voyage AI and OpenAI may process a request outside the United States (section 13). We do not sell Google user data, use it for advertising, or transfer it to anyone else except as the law requires or as part of a sale or merger of Rivyt (section 4).

No AI training

Rivyt does not use Google Workspace API data to develop, improve, or train non-personalized AI and/or ML models. Google user data is used only to answer the questions of people in the company that connected it, never to build a model for anyone else.

Limited Use

Rivyt's use and transfer to any other app of information received from Google APIs will adhere to Google API Services User Data Policy, including the Limited Use requirements.

Google API Services User Data Policy

Who at Rivyt can read it

Rivyt staff do not read Google user data except: with the agreement of the person or company concerned, for specific data, for example to give support they asked for; to investigate a security problem or abuse; or to comply with the law.

How we protect it

• Encryption in transit everywhere, and encryption at rest by our providers.
• Google access tokens encrypted again by Rivyt.
• The Drive connection is bound to the person who started it.
• Each company’s data is kept separate in the application and again in the database, and each workspace has its own search area.
• Documents placed in AI prompts are marked as data, not instructions, to resist prompt injection.
• AI tools are read-only and scoped to one workspace (section 12).

How long we keep it, and how to delete it

• Files from Drive, and everything Rivyt made from them, are kept while your company uses Rivyt, and deleted at the end of its contract as section 2 describes.
• Disconnecting Drive revokes Rivyt’s access at Google and deletes the stored access tokens. Files already synced stay in Rivyt unless a workspace admin chooses to delete them, when disconnecting or later from the Google Drive panel on the Documents page.
• Deleting a synced file, or all of them, removes the stored files, their search index entries and the text kept from them, and withdraws what Rivyt learned from them, so it is no longer used (only a record of the withdrawal is kept).
• If a file is deleted or trashed in Drive, Rivyt removes its search index entries.
• Deleting a workspace deletes all of these for every document in it.
• Removing Rivyt in your Google Account’s security settings stops syncing but deletes nothing in Rivyt; ask your admin, or privacy@getrivyt.com, to delete what was synced.
• Answers already given that quoted a file stay in question history until it is deleted (section 2).

Your company can export it

A company admin can download an export of the workspaces they belong to, including the text Rivyt read from each document, the original files, question history and the knowledge graph.

6. Voice

Voice is off until you turn it on, and Rivyt works fully by typing. If you turn it on, OpenAI handles it:

• What you say. Your recording is sent to OpenAI to be turned into text. Nothing is recorded until you tap the microphone, and recording stops when you tap it again. While you speak, the recording so far is sent every couple of seconds so your words can appear as you talk.
• What Rivyt answers. If spoken replies are on for your workspace, the text of the answer is sent to OpenAI to be read back. Your documents are never sent whole, only the sentences Fox would speak, which are written from your documents and can quote them.
• Not used for training. OpenAI does not train its models on data sent through its API. That applies to your audio and to the answers read back.
• How long it is kept. Rivyt does not keep your audio at all: it is transcribed and discarded. OpenAI holds it for up to 30 days for abuse monitoring, then deletes it.

No voiceprints. We do not create voiceprints, and we do not use your voice to identify you or anyone else. The text of what you said becomes your question, and is kept as question history (section 2).

Only record yourself. Ask before recording anyone else. Some states require everyone in a conversation to agree before it is recorded. Your company may have its own rules for recording on its sites.

You can turn voice on or off at any time in Settings. The full details are in Voice and your privacy.

6A. Inspections

If your company uses inspections:

• What is kept. The text of what the technician says (the recording is turned into text by OpenAI, as for voice, and discarded), the ratings and notes, the photos and short videos they take, who signed the inspection and when, and the vehicle’s details. A shop customer’s contact details are kept only for the shop’s own list.
• AI. The text of the inspection is sent to Anthropic to map what was said to the checklist and to word the customer’s report. Signed findings can be added to your company’s knowledge graph, without the customer’s details.
• Photos lose their location. Rivyt removes the hidden details a phone writes into a photo, including where it was taken, before storing it.
• The vehicle identification number is sent to the US National Highway Traffic Safety Administration’s free public decoder to look up the make, model and year. Only the VIN is sent.
• The customer’s link shows only the signed report: the findings, photos the technician chose to share, the vehicle’s year, make, model, mileage and the last six characters of its VIN. It does not show the customer’s name or phone number, the technician’s own notes, or what was said. Links expire, the shop can stop one at any time, and they are hidden from search engines.

7. How AI is used

• What the AI does. Rivyt uses AI models from Anthropic to write answers from your company’s documents and to build a knowledge graph from them, and models from Voyage AI to search and rank passages. Each answer cites the passages it came from.
• No training on your data. Neither we nor our providers use your content, your questions or your voice to train AI models. Our AI providers’ API terms say they do not train on API data, and where a provider offers a setting to opt out of training, we have opted out.
• Answers can be wrong. Check the cited source before you act on an answer. Rivyt never replaces your company’s safety procedures, lockout and tagout, or any controlled procedure.
• Decisions stay with people. When Rivyt compares options (for example, repair or replace), the arithmetic is done by code, every figure cites its source, and Rivyt does not recommend or choose. Rivyt does not rank, score or make decisions about people.
• Knowledge from people. Knowledge you share in an interview is attributed to your role, not shown with your name, and reviewed by your company’s admins before it is used.

8. Your rights and how to use them

Wherever you live in the United States, you can ask us to:

• Know: tell you what personal information we hold about you, where it came from, why we use it, and who received it;
• Access and portability: give you a copy of it, in a common format you can take elsewhere;
• Correct it if it is wrong;
• Delete it;
• Opt out of any sale or sharing, of targeted advertising, and of profiling that has legal or similarly significant effects (section 10). We will confirm your opt out in writing;
• Limit the use of sensitive personal information. We already use it only to provide the service (section 11), and we will still confirm your request;
• Appeal if we refuse a request (below).

Do it yourself. You can edit your name, job title and profile on your profile page, switch answer tailoring off, and turn voice off in Settings at any time.

Ask us. Email privacy@getrivyt.com. Say what you want and which company invited you to Rivyt.

How we confirm it is you. We reply to the email address on your Rivyt account and ask you to confirm the request from it. We do not ask for ID documents. If we cannot confirm who you are, we tell you why and what would work instead. We use what you give us for verification only to verify you.

Someone acting for you. An authorised agent can make a request for you with your signed written permission, or under a power of attorney. We may ask you to confirm your identity with us directly and to confirm that you gave the agent permission.

When we answer. We acknowledge your request within 10 business days, and answer within 45 calendar days. If we need longer, we tell you why within those 45 days, and take no more than 45 more. We act on an opt out request within 15 business days.

What deletion does not reach at once. Backups age out on their normal cycle. Audit records of access to your account are kept, without your name, while your company uses Rivyt. OpenAI keeps voice data under its own terms (section 6). We tell you about each of these when we answer.

If we say no. We tell you why. You can appeal by replying to our answer with “Appeal” in the subject line. We review the decision again and answer within 45 days. If you are not satisfied, you can contact your state attorney general; we give you the link in our answer.

We will not treat you differently for using any of these rights. We will not deny you the service, charge you a different price, or give you a different quality of service because you used them. We offer no financial incentives for personal information.

California’s Shine the Light law. We do not disclose personal information to third parties for their own direct marketing.

9. If you use Rivyt through your employer

Most of what Rivyt holds about you, it holds for your company. Your company’s documents, and the questions and answers in its workspaces, are its records, and it decides what happens to them. For those, Rivyt is your company’s service provider, and your company is responsible for answering your privacy requests.

• Send requests about your company’s content to your employer. If you send them to us, we pass them to your company, tell you we did, and help your company respond.
• Your account and profile are about you. We act on requests about them directly, and tell your company’s admin.
• Your questions and answers are about you and held for your company. We act on them, and tell your company’s admin.
• When you leave your company, your company removes your account. Your profile and sign in are deleted. Your company may also delete your question history, or keep it. Knowledge you shared stays with your company, attributed to your role, with no link to you.
• Workplace monitoring notices. Some states require employers to tell employees about electronic monitoring. Because your company’s admins can see question history, we give our customers a notice to use.

10. Selling, sharing and Global Privacy Control

We do not sell personal information. Nothing in the Rivyt app, and nothing about your company’s content, your account or your questions, is ever shared for advertising. The only advertising tags we use are the ones on our ad landing pages described in section 2A, which tell the advertising platform that showed you our ad that you visited. We do not knowingly sell or share the personal information of anyone under 16.

Global Privacy Control. If your browser sends a Global Privacy Control signal, we treat it as a valid request to opt out of sale and sharing for that browser: our ad landing pages do not load an advertising tag. You can also opt out by emailing privacy@getrivyt.com.

11. Sensitive personal information

The only sensitive personal information we collect is your account login (email, password and two factor authentication). We use it only to sign you in and keep your account secure, which is using it to provide the service you asked for. We do not use sensitive personal information to infer characteristics about you. Because of that, there is no “Limit the use of my sensitive personal information” link.

12. Security

• Everything is encrypted in transit, and stored data is encrypted at rest by our providers.
• Everyone signs in with two factor authentication, and sensitive actions ask for it again.
• Passwords found in known data breaches are refused, and changing your password needs your current one.
• Each company’s data is kept separate from every other company’s, in the application and again in the database.
• Rivyt staff access to customer content needs a stated reason.
• AI tools are read-only and scoped to one workspace, and nothing is written into another system until a person approves it.

No system is perfectly secure. If a breach affected your information, we would tell your company without undue delay and within 72 hours of confirming it, and tell you, and any regulator, as the law requires. To report a security problem, email security@getrivyt.com. More is on our security page.

13. Where your data is stored and processed

Rivyt is offered to companies in the United States, and stores your data in the United States: our database, file storage and search index (Supabase and Pinecone) are in the United States, as are our error reports, emails and feedback form (Sentry, Resend and Formspree).

Some processing can happen elsewhere. The AI providers we use may process a request outside the United States: Anthropic’s models may run in any region where Anthropic operates, and Voyage AI and OpenAI do not commit to a processing region for our account. Connections to our website, app and servers pass through networks that handle each connection at the location nearest the user. If your company has staff outside the United States who use Rivyt, its contract with us covers how their data is protected.

14. Children

Rivyt is not intended for anyone under 18, and is available only by invitation from a company. We do not knowingly collect information from children under 13, or from anyone under 18. If we learn that we have, we delete it.

15. Do Not Track

The Rivyt app does not track you across other sites or services, so a browser’s Do Not Track setting has nothing to switch off there. Our ad landing pages honour Global Privacy Control as section 10 says.

16. Changes to this policy

We review this policy at least once every 12 months, and whenever what we do changes. When we change it, we update the date at the top. If a change is significant, we tell you in the app and tell your company’s admins before it takes effect. We will not use personal information in a materially different way from what this policy said when we collected it without telling you first and, where the law requires, asking your consent.

This policy is available in an accessible format on request.

17. Contact

• Privacy questions and requests: privacy@getrivyt.com
• Security problems: security@getrivyt.com
• Everything else: hello@getrivyt.com

Naveiyra Holdings, Inc. dba Rivyt, a Delaware corporation, United States.