Privacy Policy
Last updated
Concept project. Dummy Waiter is a concept and portfolio project built by an independent developer. It is not a commercial service. No hardware is for sale, no payments are taken, and nothing on this site is an offer to sell. This page describes how the demonstration site and working prototype actually behave.
This policy explains what personal information Dummy Waiter collects, why, who handles it, how long it is kept, and the choices you have. It describes what the code does today, in plain language, rather than what a commercial product might promise.
On this page
In short
- Dummy Waiter is a portfolio project, not a business. There is no advertising, no analytics, and no sale of personal information.
- If you join the waitlist we keep your email address and role. If you sign in to the prototype we keep your account, restaurant, staff, menu, and order data.
- Speech recorded in the prototype is turned into text by AI services. The audio itself is not saved by Dummy Waiter. The transcript is saved with the order.
- A restaurant that records conversations is responsible for telling guests and for getting consent where the law requires it.
- Only essential cookies and browser storage are used. See the Cookie Policy.
- You can ask to see, correct, or delete your data at any time by emailing hello@dummywaiter.com.
1.Who we are and what this covers
Dummy Waiter is a concept and portfolio project operated by an independent developer based in Providence, Rhode Island, United States. In this policy, “we”, “us”, and “Dummy Waiter” mean that developer. You can reach us at hello@dummywaiter.com.
For the waitlist, sign-in, bug report, correspondence, and security log data described below, we decide why and how the data is used, so we are the “controller” (GDPR and UK GDPR) or “business” (California law) for it.
For restaurant data that a restaurant enters into the prototype (menus, staff, orders, transcripts, and the speech of guests and staff), the restaurant decides why and how it is used. We act as its service provider (“processor”) and handle that data on its behalf. Each restaurant is responsible for its own notices to staff and guests. See Audio recording and transcription.
This policy covers the public website, the no-login demo, the sign-in dashboard prototype, and the backend that transcribes and parses recordings. It does not cover other sites we link to, such as GitHub.
2.Information we collect
We collect only what the features need. The table below lists each category, what it includes, and where it comes from.
| Category | What it includes | Where it comes from |
|---|---|---|
| Waitlist | Email address, your role (restaurant owner, manager, waiter, or other), an automatically generated referral code, and the time you signed up. | You, through the waitlist form. |
| Sign-in and account | Email address. A password, only if you choose to set one (stored by our authentication provider as a hash, never visible to us). Your name and profile details if you use Google sign-in. An internal account ID. Session, sign-in time, and IP address records kept by the authentication provider. Staff who join a restaurant with an access code get an anonymous account with no email address. | You, Google (if you choose Google sign-in), and our authentication provider. |
| Restaurant and staff | Restaurant name. Staff names and roles. Each staff member’s PIN of 2 to 4 digits, stored only as a bcrypt hash. Restaurant access codes. Registered device records. Shift permissions and display preferences. | The restaurant owner and staff. |
| Operations | Menu categories, items, prices, and descriptions. Floor plans and tables. Orders, including table, staff member, items, modifications, allergy notes, prices, status, and timestamps. Shift summaries and analytics calculated from them. | The restaurant, and the prototype as staff use it. |
| Voice and transcripts | Audio of an order conversation when a staff member starts a recording (processed but not stored by us). The text transcript, an AI-corrected transcript, and the parsed items, saved with the order. | Staff who record, and any guests or bystanders whose voices are within range. |
| Menu uploads | PDF or photo files of a menu that a user chooses to import. Files are read in memory and are not saved by us. | The user who uploads them. |
| Bug reports and feedback | Subject, description, severity, an optional screenshot, the reporter’s name and role, the restaurant name, browser and device details (user agent), and the time. | The user who submits a report. |
| Technical and security data | IP address, browser type, pages requested, and timestamps in server logs. Rate-limit counters. Application logs, which can include route names and internal account IDs. | Collected automatically from your device and our servers. |
| Browser storage | A few cookies and storage items needed for sign-in and the interface. The full list is in the Cookie Policy. | Your browser. |
| Correspondence | Messages you send us, for example by email. | You. |
We do not knowingly collect payment card details, government identification numbers, precise location, or biometric identifiers, and the prototype does not take payments. The prototype does not create voiceprints or identify people by their voice. An order’s allergy note can reveal health information, so please enter only what an order needs.
The no-login demo runs on built-in sample data. Its recording, transcription, and menu import steps are simulated with canned results, so the demo does not use your microphone or send audio or files to any AI provider. It sets a single cookie to remember that you chose it (see the Cookie Policy). Beyond that and standard server logs, we do not collect personal information from people who browse the demo.
3.How we use information and our legal bases
If you are in the European Economic Area, the United Kingdom, or Switzerland, the table also names the legal basis we rely on.
| Purpose | Information used | Legal basis (GDPR and UK GDPR) |
|---|---|---|
| Run the waitlist and send a one-time welcome email with your referral code | Waitlist data | Consent, given when you submit the form. You can withdraw it at any time. |
| Create and secure accounts, sign you in, and run the prototype features you use | Account, restaurant, staff, and operations data | Performance of a contract (our Terms of Service). For restaurant data, we act on the restaurant’s instructions as its processor. |
| Transcribe recorded audio and parse orders with AI | Audio, transcripts, menu data | On the restaurant’s instructions as its processor. The restaurant is responsible for its own legal basis and notices. |
| Keep the service secure, prevent abuse, apply rate limits, and fix problems | Technical and security data | Legitimate interests in running a secure and reliable service. |
| Respond to bug reports, questions, and requests | Bug reports and correspondence | Legitimate interests, and contract where you are an account holder. |
| Comply with the law and handle legal claims | Any of the above, as needed | Legal obligation, and legitimate interests. |
| Tell you about material changes, or if the project is wound down | Email address | Legitimate interests in keeping people we hold data about informed. |
What we do not do
- We do not sell personal information.
- We do not share personal information for cross-context behavioral advertising, and we do not use it for advertising of any kind.
- We do not run analytics, advertising, or cross-site tracking.
- We do not send marketing newsletters.
- We do not use your data to train AI models. Transcripts and menu text go to AI providers only to produce a result for you, and those providers state in their API terms that they do not train on that data.
- We do not make decisions with legal or similarly significant effects about anyone by automated means. AI output is a suggestion, and a staff member reviews and approves each order before it is sent to the kitchen.
4.Audio recording and transcription
Dummy Waiter records speech, so this is the section to read if you are a restaurant, a server, or a guest. The wearable clip on the landing page is a concept and no hardware exists to buy. The working prototype records through the microphone of the device and browser that a staff member is using.
How recording works in the prototype
- Recording starts only when a signed-in staff member taps record on the Orders screen, and only after the browser has asked for and received microphone permission. It stops when they tap stop. The prototype does not listen continuously or in the background.
- The audio is sent over an encrypted connection to the Dummy Waiter backend, which is hosted on Railway.
- The backend forwards the audio to Groq, which runs the Whisper large v3 speech-to-text model and returns a transcript.
- The transcript, together with the restaurant’s available menu items, is sent to Anthropic (Claude) to correct obvious transcription errors and pick out the ordered items. If that step is unavailable, a rule-based parser on our own server is used instead.
- The result appears on the staff member’s screen. Nothing is sent to the kitchen until a staff member reviews and approves it.
What is kept
- Audio. Our code does not write audio to a database, file, or storage bucket. It exists in memory in the browser and on the backend while it is processed. Groq states that by default it does not retain inference inputs or outputs, but that it may log them temporarily, for up to 30 days, to troubleshoot errors or investigate abuse.
- Transcripts. The original transcript, the corrected transcript, and the parsed items are saved with the order, but only when staff submit the order. They are visible to that restaurant’s signed-in staff and are kept until the order or restaurant account is deleted (see Retention).
- AI provider copies. Anthropic states in its published API data retention policy that inputs and outputs are deleted from its systems within 30 days by default, with limited exceptions, and are not used for training without permission.
Who may be recorded
Any speech within range of the microphone can be captured, including guests, colleagues, and people at nearby tables. We do not identify speakers by their voice or keep audio for training.
Notice and consent are the restaurant’s responsibility
Recording a conversation is regulated. In the United States, federal law and many states allow a recording when one participant consents. Others, including California, Florida, Illinois, Maryland, Massachusetts, Pennsylvania, and Washington, generally require every participant to consent. In the EU and UK, a recording of an identifiable person is processing of personal data that needs a lawful basis and clear notice. Rules differ by country and change over time.
A restaurant that uses the recording feature is responsible for giving guests and staff clear notice (for example, signage or a spoken notice from the server), for getting consent where the law requires it, for having a lawful basis under privacy law, and for stopping a recording when asked. Dummy Waiter cannot do this for you and does not give legal advice. The Terms of Service set out the same responsibility and include sample notice wording.
Please do not record real guests in this prototype
This is a demonstration. Unless you have met the notice and consent requirements above, use test speech or the sample data in the demo instead of recording real guests.
If you are a guest and believe your voice was captured by this prototype, speak to the restaurant first. You can also email hello@dummywaiter.com and we will help the restaurant find and delete the related transcript.
7.How long we keep information
| Information | How long we keep it |
|---|---|
| Waitlist data | Until you ask us to remove it, or the project is retired, whichever comes first. We review the list at least once a year. |
| Account and sign-in data | While your account exists. Deleted within 30 days of a verified request. Prototype accounts with no sign-in for 12 months may be deleted. |
| Restaurant, staff, menu, floor plan, orders, transcripts, and shift summaries | Until the restaurant deletes them or asks us to delete the account, or the inactive-account rule above applies. |
| Audio | Not stored by us. Held in memory only while it is processed. Groq may keep temporary logs for up to 30 days in limited cases. |
| Text sent to Anthropic for parsing | Up to 30 days under Anthropic’s default API retention, with limited exceptions. |
| Bug reports and screenshots | Up to 24 months, or until you ask us to remove them. |
| Server and application logs | Kept by our hosting providers under their default settings, and used only for debugging and security. |
| Email correspondence | Up to 24 months. |
| Backups | Deleted data can remain in our database provider’s backups for a limited period before it is overwritten. |
There is no self-service account deletion in the prototype yet. Email hello@dummywaiter.com and we will delete your data by hand. We may keep information longer where the law requires it, or where we need it to resolve a dispute or enforce our terms.
8.Your privacy rights
Everyone
Wherever you live, you can ask us to tell you what we hold about you and give you a copy, correct it, delete it, restrict or stop using it, or give it to you in a portable format. If we rely on your consent, you can withdraw it at any time. We apply these choices to everyone, not only where the law requires them.
European Economic Area, United Kingdom, and Switzerland
Under the GDPR and UK GDPR you have the rights of access, rectification, erasure, restriction of processing, data portability, and objection (including to processing based on legitimate interests). Withdrawing consent does not affect processing that happened before you withdrew it. You may also complain to your data protection authority. In the UK that is the Information Commissioner’s Office (opens in a new tab), and in the EU you can find your authority in the European Data Protection Board’s member list (opens in a new tab). We would appreciate the chance to put things right first.
This project is not directed at people in the EU or UK and has not appointed an EU or UK representative. The rights above still apply if you contact us.
California and other US states
This is a small, non-commercial project and may fall below the thresholds that make the California Consumer Privacy Act (as amended by the CPRA) apply. We follow its principles anyway. California residents, and residents of other states with similar laws, can request to know and access, delete, or correct their personal information, can opt out of its sale or sharing, can ask us to limit the use of sensitive personal information, and will not be treated worse for exercising these rights.
In the past 12 months we have not sold or shared personal information as those terms are defined in California law. We use sensitive personal information (account log-in credentials, which are stored as hashes, and any health details a user puts in an allergy note) only to provide the service. The categories we collect are:
| Category | Examples | Disclosed for a business purpose to | Sold or shared |
|---|---|---|---|
| Identifiers | Name, email address, IP address, account ID, browser details | Supabase, Vercel, Resend, Railway | No |
| Customer records | Staff names, restaurant name | Supabase | No |
| Commercial information | Orders and menu data entered by restaurants | Supabase, Railway, Anthropic | No |
| Internet or network activity | Server logs, bug report device details | Vercel, Supabase, Railway | No |
| Audio and electronic information | Recorded speech (processed, not stored) and transcripts | Railway, Groq, Anthropic, Supabase | No |
| Professional information | Role (owner, manager, waiter) | Supabase | No |
| Sensitive personal information | Account log-in credentials (hashed), incidental health details in allergy notes | Supabase | No |
We do not collect biometric information, precise geolocation, or education records.
How to make a request
- Email hello@dummywaiter.com with the subject “Privacy request”. Tell us the email address you used and what you would like us to do.
- We may ask you to confirm your identity, for example by replying from the same email address, so that we do not give your data to someone else.
- We will respond within 30 days. Where the law allows and the request is complex, we may extend that by up to 60 more days and will tell you why.
- Requests are free. If we decline all or part of a request, we will explain why and how to appeal by replying to us.
- An authorized agent can make a request for you with your written permission.
- If a restaurant entered the data, we may need to pass your request to that restaurant, or act on its instructions.
9.Children
Dummy Waiter is a tool for restaurant owners and staff and is intended for people aged 18 and over. It is not directed at children, and we do not knowingly collect personal information from anyone under 16. If you believe a child has given us personal information, email hello@dummywaiter.com and we will delete it.
10.International data transfers
We are based in the United States and our providers process data mainly in the United States. If you are in the EEA, the UK, or Switzerland, your information will be transferred to, and stored in, the United States and possibly other countries. The region where our database is hosted is set in our provider account and may be outside your country.
Where the law requires safeguards for these transfers, our providers rely on mechanisms such as the European Commission’s Standard Contractual Clauses, the UK International Data Transfer Addendum, and, for providers that are certified, the EU-U.S. Data Privacy Framework. Email hello@dummywaiter.com if you would like more detail.
11.Security
- Data is encrypted in transit over HTTPS (with HSTS), and our database provider encrypts stored data.
- Row-level security in the database limits each restaurant’s data to its own signed-in devices. Privileged database keys are used only in server-side code.
- PINs are stored only as bcrypt hashes. Passwords are handled by our authentication provider. Repeated wrong PIN or access code attempts are rate limited and temporarily locked.
- Staff sessions lock after 10 minutes of inactivity and need a PIN to reopen.
- Text that users submit is sanitized before it is stored, and standard browser security headers are sent. Microphone access is limited to this site.
No system is perfectly secure. This prototype has not had an independent security audit or a certification such as SOC 2, so please do not store sensitive or regulated data in it. Screenshots attached to bug reports are saved in a storage bucket whose files can be opened by anyone who has the exact link, so do not include personal or confidential information in them. If a breach affects your personal information, we will tell you and the relevant authorities as the law requires.
12.Changes to this policy
We may update this policy as the project changes. When we do, we will change the date at the top of the page, and for material changes we will take reasonable steps to tell you, such as a notice on the site or an email to account holders. Using Dummy Waiter after a change means you accept the updated policy. If you do not, please stop using it and ask us to delete your data.
13.Contact
Questions, requests, and complaints about privacy: hello@dummywaiter.com. This project has no registered office. If you need to send a formal notice by post, email us first and we will give you an address.
See also the Terms of Service, the Cookie Policy, and the Accessibility Statement.