Last updated: 3 November 2026
This policy explains what personal data Made Louder collects, why, who else handles it, how long we keep it, and what rights you have. Made Louder is a product operated by Pixebyte Ltd, a company registered in England and Wales (company number 13869005), registered office 28 Maes Sarn Wen, Four Crosses, Llanymynech, Powys, United Kingdom, SY22 6NT. Pixebyte Ltd is the data controller. You can reach us about privacy at privacy@madelouder.com.
Made Louder is a writing and publishing assistant for founders. You talk to your assistant on WhatsApp, and use a web console at app.madelouder.com to give it your material, review its work and connect your accounts. It learns how you write and drafts posts in your voice for you to approve. If you film yourself, it makes your take into a short video, also for you to approve. We publish what you approve to your LinkedIn for you once you have connected your account. Until you do, nothing is posted for you, and you post the pieces you approve yourself.
Made Louder runs on the Hundredloop service, which Pixebyte Ltd also operates, so some of the addresses Made Louder uses are on hundredloop.ai. Today these are the address WhatsApp delivers your messages to, the address LinkedIn returns you to when you connect your account, and media.hundredloop.ai, where files we make or store for you are kept. Once the features that use them are switched on, they will also include the addresses Facebook uses when you connect or remove a Page and the unsubscribe links in our emails. If you use Made Louder, this policy is the one that applies to you, including at those addresses.
The assistant you talk to is an artificial intelligence, not a person, whatever name you have given it. Its replies and your drafts are written by AI models. If you need help from us, start a message with @support, or email hello@madelouder.com. An AI may answer first, from guides we have written, and it says so when it does. If that does not sort it, write @support again and one of us will look at it.
| When | What | Why (lawful basis) |
|---|---|---|
| We set up your account | Your name, your business's name, the name of your brand, your time zone, your email address if you give us one, and the name you choose for your assistant. Your WhatsApp phone number is linked to your account, either when we set it up or when you first message us from the invitation we send you. Your number is how we recognise you | Contract |
| You open the console | There are no passwords. You get in with single-use links that your assistant sends you in WhatsApp; an unused link expires after seven days. Opening one starts a session that your browser keeps. We store only a scrambled (hashed) form of each link and session secret, never the secret itself | Contract |
| You message your assistant | What you write, when you sent it, and what your assistant replies. Your messages are free-form, so they hold whatever you choose to tell us about yourself and your work | Contract |
| You write to support | What you write with @support, when you wrote it, what we reply, and a record of the request: where it has got to, why it came to us, how often you wrote, a short summary, and any notes our staff add | Contract |
| You send a voice note | We fetch the recording from WhatsApp, keep it with your account, and send it to OpenAI to be turned into text. The transcript is kept as part of your conversation. We keep the recording so that your own voice can be the soundtrack of a reel, which is only ever published with your approval. We do not use your voice to identify you, and we do not use it to make a synthetic copy of your voice. Voice notes are optional; you can always type | Contract: transcribing a voice note is how your assistant understands what you said, and keeping it is how your own voice can be used in your reels |
| You give us the addresses of your work | The text of the public pages at the addresses you give us, with each page's address, title and when we read it, and what we work out from them: how you write, what you stand for, and the facts about you that they support. We read only the pages at those addresses. We do not follow their links, and we do not search the web for you | Contract. You confirm you have the right to point us at them |
| You add documents in the console | The text of the documents you add: PDF and Word (.docx) documents, and text files such as .txt, .md, .csv and .html. We keep each document's name and its words, not the file. They may contain other people's details | Contract. You confirm you have the right to share them |
| You film in the console | The video and sound you record, which will usually show your face and voice. The console holds each take on your device until it has uploaded, then clears it from your device. We keep each take with your account, with the copies we make of it to work from. We do not use your face or voice to identify you, and we do not use what you film to make a synthetic copy of you or your voice | Contract |
| You ask for a video of a post | The script we write from your post for you to read, a transcript of each take with the time each word was said, the cut we make from that, and each finished video with its captions. Software makes the cut and the captions from the transcript. A finished video goes out only after you have said yes to that video. A take and a finished video are each kept at a long, unguessable web address so that our software can fetch it and the console can show it, which means anyone given that address could open it | Contract |
| You connect LinkedIn | The access LinkedIn grants us (access and refresh tokens, and the permissions you approved), which we encrypt before we store it, and your LinkedIn member id, name and email as LinkedIn supplies them | Contract. We use the access only to confirm which LinkedIn account you connected and to publish what you have approved |
| We draft and publish for you | The drafts, your edits and decisions, the pictures we make to go with your posts and what we asked for when making each one, the schedule, when each post went out and its id on LinkedIn, and anything you tell us came of it, such as replies or calls booked. We do not currently read performance figures back from LinkedIn. A picture is kept at a long, unguessable web address so that WhatsApp can fetch it and the console can show it, which means anyone given that address could open it | Contract |
| You email us, or we email you | Your email address and what the message says | Legitimate interests: answering you and running your account |
| Our systems run | Operational logs and error records: what the system did and when. They show only the last four digits of a phone number and do not record the text of your messages. Logs written before 30 September 2026 can also include your full phone number, the first 120 characters of each text message you sent your assistant, and a few words from other messages, such as the name you gave your assistant. Logs are deleted after 30 days | Legitimate interests: keeping the service secure, working and fixable |
| You visit our website or console | The technical details every web request carries, such as your IP address and browser type, which our hosting provider handles to deliver the page. We run no analytics or advertising trackers | Legitimate interests: delivering the site |
Where we say "contract", we mean providing the service under the agreement you have signed with us. If that agreement is with your company rather than with you personally, we rely instead on legitimate interests: yours, your company's and ours in providing the service your company signed up for.
If you send a photo, a document or any other attachment in WhatsApp, your assistant cannot use it yet, and we do not keep it.
We do not ask for sensitive information, such as your health or your religious or political beliefs. If you choose to share it in your conversation or your material, it becomes part of what your assistant works from, and you will see any draft that uses it before that draft can be published.
What you tell us and give us may mention other people: clients, colleagues, family. We use it only to write for you, and nothing about them is published unless you approve the post it appears in. If you message our WhatsApp number without an invitation, we do not reply and we do not keep your message; our logs record that it arrived and the last four digits of the number it came from, and are deleted after 30 days.
The heart of this product is reading what you have already published and using it to write in your voice. We read only the pages at the addresses you give us and the documents you add. We do not buy data about you, scrape people who have not asked us to, or build profiles of anyone other than our own customers. Where we could not find something, we say so rather than guessing.
AI models read your material and your conversation to learn your voice, to ask you questions and to draft your posts. A model also works out when each word of a take was said, so that software can cut and caption it. We do not use your material, your conversation or what you film to train AI models. When you write to support, an AI model may read your request, a few of your recent messages and basic facts about your account, to see whether one of our written guides answers it. It can only pick a guide or pass your request to us. It does not write its own answer, and it changes nothing on your account. No decision with legal or similarly significant effects on you is made by automated means: the AI drafts, and you decide what is published.
Other customers cannot see your data. Inside Pixebyte, a small number of people can see your account, including your conversation, when they need to: to help you, to fix a fault, or to check that the service is working for you. We do not look at it for any other reason.
When you write to support, your request is kept in your account. One of us gets a WhatsApp alert with your first name, your business's name and a link to our staff portal: not your number, and not what you wrote. We read and answer it in the portal, which needs a password and an authenticator code. Each time one of us opens your conversation or changes your account there, we record who did it and when, but not what they read.
We do not sell your data. We use a small set of service providers to run the product, each under data processing terms that require it to protect your data and to use it only to provide its service to us:
| Provider | What they do | Where |
|---|---|---|
| Amazon Web Services | Hosting: our servers, database, file storage, message queues and logs, and the software that cuts and captions your videos. Our web pages are delivered through AWS's worldwide network | United Kingdom (London) |
| Anthropic | The AI models behind your assistant, the interview, the drafting, choosing what the picture for each post shows, and the check on support requests | United States |
| OpenAI | Turning voice notes into text, timing the words in your videos, and making images: pictures for your posts, and images from the spoken words of the video scripts you approve. For a video we send the sound of each take and its script, as soon as the take has uploaded and before you have seen the video. For a post's picture we send a short description of what to show, the colours chosen for you and, if you ask for a different picture, what you said about it. If we have no description we send the opening of the post instead. This can happen before you have approved the post. We do not send photos of you, or the picture from anything you film | United States |
| Meta (WhatsApp) | Carrying messages between you and your assistant, including the pictures we make for your posts, and alerts to us that a support request is waiting | Meta's own infrastructure, under the WhatsApp Business terms |
| The typefaces our website and console use. Your browser fetches them from Google Fonts, so Google receives your IP address | Google's own infrastructure | |
| Stripe | Card payments, if you pay us by card. Stripe's hosted checkout takes your card details, so they never reach our systems | Stripe's own infrastructure |
When we publish for you, LinkedIn receives the posts you approve, with their pictures, and the videos you say yes to, and handles them, and your account, under its own privacy policy. What has gone out stays on LinkedIn until you remove it there. Publishing to Facebook and Instagram is built but not live yet, and those accounts cannot be connected yet. We will switch it on once Meta has approved our access, and Meta will then receive what you approve for those accounts in the same way. We will update this page before we add a provider that receives your personal data.
Our database, stored files and logs are held in the UK, in AWS's London region. Anthropic and OpenAI process what we send them in the United States, and Meta and Google may process data outside the UK. Where personal data leaves the UK, we rely on the safeguards UK law provides for it: the UK-US data bridge where a provider is certified under it, or the UK's international data transfer addendum to the standard contractual clauses in the provider's data processing terms.
Under UK data protection law you have the right to access your data, correct it, delete it, restrict or object to how we use it, and ask for a copy in a portable form. You can download a copy yourself at any time with "Download my data" on the Accounts page in the console: one file with your details, your brands, your conversation, your support requests, what we have written for you, the material you gave us and what we learned from it, and which accounts are connected (but not the access tokens for them). Notes our staff write on a support request are not in the download. Ask at privacy@madelouder.com and we will include them. It holds the words of your videos and the addresses of the finished ones, not the video files or what you filmed: ask us and we will send you those. To exercise any of these rights, email privacy@madelouder.com and we will respond within one month. If you are unhappy with how we handle your data, you can complain to the Information Commissioner's Office (ico.org.uk), though we would rather you told us first so we can put it right.
We keep this minimal. The website sets no cookies. The console keeps you signed in using your browser's local storage rather than a cookie, and uses the same storage to remember a few settings and to hold a video take on your device until it has uploaded. This storage is strictly necessary for the console to work. We use no advertising or analytics cookies or trackers. Our pages load their typefaces from Google Fonts, as described above.
We do not send marketing emails or messages, and we do not track whether our emails are opened. Your assistant sends you service messages in WhatsApp: questions, drafts to approve, reminders and updates on your work. See "Stopping" above to switch them off. If we ever want to send you marketing, we will ask first, and every message will carry a way to unsubscribe.
Made Louder is a business product and is not intended for anyone under 18.
We will update this page when our practices change, and we will note the date at the top. For material changes we will tell you directly.
Questions about this policy or your data: privacy@madelouder.com.