Privacy policy
Last updated: 29 September 2026
ChannelDesk is a customer service inbox for businesses. It brings the messages and comments a business gets on Facebook, Instagram, WhatsApp, email and its website’s chat into one place, where its team and Dot, ChannelDesk’s AI agent, answer them. It also runs the business’s help centre and tickets. This policy explains what data that involves, what we do with it, who helps us run the service, how long we keep it and how to have it deleted.
Who we are
ChannelDesk is run by Wecover Technologies Limited, a company registered in Nigeria with the Corporate Affairs Commission (RC 1892579), of 7 Ibiyinka Olorunbe Close, Victoria Island, Lagos 100242, Nigeria (“we”, “us”). For anything about your data, email privacy@channeldesk.ai.
Our role depends on whose data it is:
- People who use ChannelDesk. For the accounts of business owners and their teams, billing, our website and our own support, we decide what is collected and why, so we are the controller.
- A business’s customers. The messages, comments, contacts and other content a business brings into ChannelDesk belong to that business. We process them only to provide ChannelDesk to it and only as it instructs, so the business is the controller and we are its processor.
If you contacted a business that uses ChannelDesk
If you messaged, commented on, emailed or chatted with a business that uses ChannelDesk, that business decides how your messages are used and how long they are kept, and its own privacy policy applies. Ask it first about your data; we will help it answer. You can also email privacy@channeldesk.ai, and we will pass your request to the business and work with it. What ChannelDesk holds on a business’s behalf is described below under From the channels a business connects.
What we collect
When you use ChannelDesk
- Your account: your name, email address and sign-in details. Sign-in is handled by Clerk. If you sign in with Google or Apple, Clerk receives your name and email address from them. You can add a photo; customers see your first name and photo in website chat.
- Workspaces and teams: the workspaces you create or join, who is invited, each person’s role, their notification settings and whether they are marked away.
- Billing: your plan, billing currency and the customer and subscription references Stripe gives us. Card details go straight to Stripe; we never see or store them.
- What you set up: the website pages you ask us to read, approved answers, saved replies, help articles, rules, workflows, Dot’s procedures, ticket types, and the apps you connect (their keys and tokens are encrypted).
- A record of what happened: an audit log of actions taken in a workspace, by a person, a rule, a workflow or Dot, and why.
- The mobile app: if you allow notifications, a push token and whether the phone is an iPhone or Android. The app uses the camera, photo library and microphone only when you choose to take or attach a picture or record a voice note, and only what you send leaves the phone. The app has no analytics or advertising.
From the channels a business connects
We receive this only when a business connects the channel, and only for that channel.
- Facebook Pages: the Page’s name and ID and an access token (encrypted). The comments on the Page’s posts and ads, the Page’s Messenger conversations, the post a comment was left on, and each person’s name, Facebook ID and profile picture as Meta provides them.
- Instagram professional accounts: the account’s name, username and IDs and an access token (encrypted). Comments on its posts and its direct messages, and each person’s name, username and profile picture.
- WhatsApp Business numbers: the phone number, its WhatsApp Business Account ID, an access token and the number’s registration PIN (both encrypted). Messages with customers, each customer’s phone number and WhatsApp profile name, message templates, delivery and read receipts, what Meta charges for messages, and the details of an ad a chat started from (its headline, text and link).
- Email: the emails a business receives at its ChannelDesk address and sends from it: names and addresses, subject, text, attachments, and whether the sender could be verified. The original email file is deleted as soon as the message is stored.
- Website chat: the messages and files a visitor sends, the name, email address and other details they choose to give, the page they are on (its address, title and the page that led there), their browser’s language and time zone, their approximate location (country and city, worked out from their IP address, which is not stored) and their kind of device (for example “Chrome on macOS”). If the business signs its visitors in, the ID it gives them. The chat keeps one item in the visitor’s browser so a conversation carries on across pages and visits; it sets no cookies.
- Files and voice notes: pictures, videos, audio and documents sent in conversations are copied to our storage, so they last as long as the conversation. Voice notes on WhatsApp are written out as text, which a business can turn off for each number.
- Customer records: each customer’s name, contact details and profile picture, the fields the business keeps about them, events the business sends (such as an order from its shop), ratings and what customers write with them, tickets, and notes the team writes.
- Help centres: the articles a business publishes, what visitors search for and ask (with no IP address or visitor ID, and with any email address typed into a search removed), and daily counts of views.
- What ChannelDesk did: the label given to a comment, any reply drafted, what Dot answered or did and why, and when a conversation was handed to a person.
When you visit channeldesk.ai
This website has no analytics, advertising or tracking scripts. It sets one cookie of its own, country, to remember which country’s prices you chose, for a year. It has our own website chat, which works as described above; what you write there comes to us. Like any website, our hosting provider, Cloudflare, processes your IP address and basic request details to deliver the page and protect it from abuse. Signing in to the app sets the cookies Clerk needs to keep you signed in. The app’s API reference loads its fonts from Google Fonts.
How we use it
- To provide ChannelDesk: bring a business’s messages and comments into its inbox, send its team’s and Dot’s replies, apply its rules and workflows, run its help centre and tickets, and report on how its customers are looked after.
- To send the emails the service needs: invitations, notifications, alerts about channels, and account and billing messages.
- To send notifications to the mobile app, if you allow them.
- To bill for paid plans.
- To keep the service secure, find and fix faults, and prevent abuse.
- To answer you when you contact us.
We do not sell data, we do not use it for advertising, and we do not use a business’s conversations for anything other than providing ChannelDesk to that business. Our legal bases are performing our contract with the business, our legitimate interest in running a secure and reliable service, complying with the law (for example, keeping billing records), and your consent for push notifications, which you can withdraw in your phone’s settings.
Data from Facebook, Instagram and WhatsApp
We use what we receive from Meta’s platforms only to provide ChannelDesk to the business that connected the Page, account or number: to show its messages and comments in its inbox, to help its team and Dot answer them, to apply its rules and workflows, and to report on them. We do not sell it, license it, use it for advertising or to build profiles, pass it to data brokers, use it to discriminate against anyone or to decide their eligibility for housing, work, insurance, credit or benefits, use it for surveillance or to re-identify anyone, train AI models on it, or use it for anything else. It goes only to the service providers listed below, which process it for us, and to apps the business itself chooses to connect. We follow Meta’s Platform Terms and Developer Policies, and a business can disconnect a channel, or remove ChannelDesk on Facebook or Instagram, at any time (see how to delete your data).
How the AI is used
Dot, ChannelDesk’s AI agent (a business can give it another name), is off until a business turns it on for a channel. It can suggest replies for the team to send, or answer customers itself. While it is off, a teammate can still ask it to suggest a reply. When it answers, it uses only what the business has given it (its approved answers, help articles, website and connected apps), it says it is an AI assistant, and the customer can ask for a person at any time. It hands the conversation to the team when it cannot answer, when the customer asks, or when the business’s rules say so. A business can review everything Dot said and did.
ChannelDesk also uses AI for other parts of the service, for example to label comments, draft replies, tell when a customer wants a person, detect the language of an email, summarise a conversation when it is handed over, suggest a ticket’s title and priority, draft an issue for a business’s bug tracker, read replies to rating requests, group what customers say in low ratings, find questions Dot could not answer, draft and test Dot’s procedures, help write help articles, read the website pages a business gives it, answer questions in its help centre, and write out voice notes.
- To do this we send the text involved (for example a customer’s message and the parts of the business’s knowledge that bear on it) to language models through OpenRouter, which passes each request to one of the hosts that run the model. We only allow hosts that do not keep the data for their own use or train on it. Some keep requests for a short time to check for abuse.
- Voice notes are written out by a speech model on Cloudflare’s own network.
- Neither we nor these providers train AI models on a business’s data or its customers’.
- Answers a business approves are kept for that business alone. Nothing one business teaches Dot is used for another.
Who processes data for us
We use these services to run ChannelDesk. Each receives only what it needs for its part, and processes it only for us.
| Service | What for | Where |
|---|---|---|
| Cloudflare | Hosting the app, this website and help centres; storing files; background work; sending and receiving email; writing out voice notes | Worldwide network; files stored in Western Europe |
| PlanetScale | The database that stores each workspace | London, UK (on Amazon Web Services) |
| Clerk | Sign-in and account security | United States |
| Stripe | Payments and subscriptions | United States and elsewhere |
| OpenRouter, and the model hosts it picks for each request | The AI features described above | United States and elsewhere |
| Expo, then Apple or Google | Delivering notifications to the mobile app, which show a customer’s name and up to 160 characters of their message | United States |
Meta is not our processor but the platform itself: when a business connects Facebook, Instagram or WhatsApp, its messages come from and go to Meta, under Meta’s own terms. The same is true of the apps a business chooses to connect, such as Slack, Shopify, Stripe, HubSpot, WooCommerce, Cal.com, Calendly, Acuity Scheduling, Jira, Linear, GitHub, its own systems or an AI tool server. Once connected, we send them or fetch from them only what the business set up, for example looking up a customer’s order by their email address, or posting a conversation into a Slack channel.
Where data goes
We are based in Nigeria. The database is in London, files are stored in Western Europe, and some of the services above work from the United States and elsewhere. We transfer data abroad under the safeguards the law requires: to countries whose protection is recognised as adequate, or under contracts that protect it, such as the Standard Contractual Clauses and, for data from the UK, the UK International Data Transfer Addendum. Those clauses also cover data that businesses in the UK and Europe send to us, as our terms set out.
How long we keep it
- Messages and comments, their files, and everything attached to them (labels, drafts and transcripts) are deleted automatically once they are older than the workspace’s retention setting: 30, 90, 180 or 365 days. The default is 180 days. A conversation and a customer record left with nothing in them go too. A comment deleted on Facebook or Instagram, or a message unsent there, loses its words here as soon as Meta tells us. When an Instagram account asks Meta for its data to be deleted, what ChannelDesk received from it is deleted straight away.
- Kept until the workspace is deleted: customer records with an email address or the business’s own ID, tickets, and the audit log, so a business keeps its customer list and can see what happened and why.
- Help centre searches and questions go with the retention setting, and after 90 days at most.
- Shorter periods: the original file of an incoming email is deleted once the message is stored (after 7 days if it never is). Meta’s notifications of new messages and comments lose their words once processed and are deleted after 30 days. The addresses a workspace emails each day are kept for two days, to hold it to its daily limit. Groupings of low ratings go after 30 days. Read notifications go after 90 days and unread ones after a year, or sooner with the retention setting. A phone’s push token is forgotten when you sign out, or after 60 days unused.
- Disconnecting a channel discards its access token straight away and stops new messages arriving. What already arrived stays until the retention setting removes it.
- Deleting a workspace cancels its subscription, disconnects every channel, deletes every access token and connected app’s key, and turns off its API keys, straight away. After 7 days the workspace and everything in it, files included, are permanently deleted.
- Deleting your account removes your sign-in and your place in every team straight away, and deletes any workspace no one else is in. What you wrote stays with your team, without your name.
- Billing records are kept as long as tax law requires. Our service logs are designed to leave out message text, names, email addresses and tokens, and are kept for a few days. Deleted data can remain in our database provider’s backups for a short time until they are replaced.
How to delete your data sets out each way to do it.
Your rights
Under the Nigeria Data Protection Act 2023, and the UK and EU General Data Protection Regulation where they apply to you, you can ask for a copy of the data we hold about you, ask us to correct or delete it, ask us to restrict or stop using it, ask for it in a portable format, and withdraw consent you gave. Email privacy@channeldesk.ai and we will answer within 30 days. If your data came to us through a business that uses ChannelDesk, we will pass your request to that business and help it answer.
Dot’s answers are not decisions with legal or similarly significant effects on anyone, and a customer can always ask for a person instead. If you are unhappy with how we handle your data, you can complain to the Nigeria Data Protection Commission at ndpc.gov.ng, or to the data protection authority where you live, such as the UK’s Information Commissioner’s Office at ico.org.uk.
Security
Data travels encrypted between browsers and phones, our services and Meta. Access tokens, keys and PINs are encrypted before they are stored. Links to files are signed and expire within hours. Each workspace is open only to the people in it, with what each can do set by their role. The part of ChannelDesk that serves requests can read and write data but cannot change how the database is structured. If a breach puts your data at risk, we will tell the businesses affected and the authorities as the law requires.
Who ChannelDesk is for
ChannelDesk is a tool for businesses, used by adults. It is not meant for children, and we do not knowingly collect their data for our own purposes.
Changes to this policy
If we change what we collect or how we use it, we will update this page and its date. If a change matters, we will tell account owners by email before it takes effect.
Contact
Privacy questions and requests: privacy@channeldesk.ai. Everything else: support@channeldesk.ai. Post: Wecover Technologies Limited, 7 Ibiyinka Olorunbe Close, Victoria Island, Lagos 100242, Nigeria.