Privacy policy
01Who we are and what this covers
Canonical is a pilot service operated by brand.co (“we”, “us”). It connects information across the work tools a team already uses and tells the right person about a connection before it becomes a problem.
This policy explains what Canonical collects, how it is used and the choices you have. It covers canonical.5pm.ai, including the public pages and the demo console.
Canonical is invite-only and is meant to be used at work by people who have been given access. It is not directed at children.
02Information we collect
Your account. You sign in with Google. From Google we receive your name, your email address and your Google account identifier. We never receive your Google password, and we do not keep the sign-in tokens Google issues.
Confirmations you give. Before you connect a source, you confirm that each client you work with has agreed to its communications being processed. We record who confirmed, for which client, when, and the exact wording shown.
A record of actions. We keep an audit record of actions such as signing in, connecting a source and disconnecting it, and technical measurements such as request counts, sizes, timings and errors. Your network address is used to limit abusive traffic.
Cookies. We set one cookie to keep you signed in, which lasts up to 24 hours, and short-lived cookies that protect the sign-in and connect steps, which last about ten minutes. We do not use advertising or analytics cookies. Your theme preference is stored in your browser.
Fonts. The pages load their fonts from Google Fonts, so your browser contacts Google to fetch them.
The demo console. The demo uses invented sample data only. What you do in it is stored in your own browser and is not sent to us.
03Connected sources
You choose which sources to connect: Slack, Gmail, Google Meet, Asana and Figma. Sources become available as the pilot progresses. Each connection is yours, and Canonical reads only what your own account can already see. We ask for read-only access wherever the provider offers it. The one thing Canonical writes is the Slack messages it sends as itself.
What is read depends on the source. Slack: messages in the channels and direct messages you are part of, and who is in those conversations. Gmail: your messages. Google Meet: transcripts and meeting notes for your meetings, where they exist. Asana: the tasks, comments and projects you can see. Figma: comments, file names and version labels for the files you can see, not the designs themselves.
We store a copy of that content, items we derive from it such as commitments, decisions and open questions, and the credentials that let us read the source. Credentials are encrypted.
Content from your sources includes messages written by colleagues and clients who are not Canonical users. We handle it only for the purposes described in this policy.
04How we use information
To provide the service: to find connections between pieces of work, to work out who should be told, and to tell them, by direct message in Slack or in an internal channel.
Connected content is processed by AI models to extract and link information. We do not use connected content to train AI models.
We do not sell personal information and we do not use it for advertising.
Working profiles. If you opt in separately, Canonical builds a profile describing how you work, such as typical reply times and preferred channels. It describes behaviour, not personality, mood or health. You can see your own profile and dispute what it says. Profiles are not for decisions about hiring, pay, promotion or discipline.
To keep the service secure, to investigate faults and to measure how the system performs.
05Who sees what
Inside your organisation, Canonical only shows a person content they could already see at the source. When a connection involves something only you can see, other people are told that you have relevant context and the general category it falls under, not the content itself.
Administrators manage who has access and see how the system is performing. They cannot read your connected content through Canonical.
Canonical does not send your information to clients.
Service providers. Google Cloud hosts the service, stores its data and provides AI processing. Cloudflare delivers the site. Slack delivers the messages Canonical sends. These providers act on our instructions.
We may disclose information where the law requires it.
06Retention and deletion
You can disconnect a source at any time. When you do, we delete our access to it and ask the provider to revoke it.
You can ask us to delete the content collected through your connections, or all the information we hold about you. Deleting content can also remove items that were derived from it.
Fixed retention periods have not been set for the pilot. We keep information for as long as the pilot runs, unless you ask us to delete it sooner. If your access to Canonical is removed, your connections are removed with it.
07Security
Information is encrypted in transit. Credentials for your sources are encrypted when stored. Access to the service is limited to invited people, and actions are recorded in an audit log. No system is perfectly secure, and we cannot promise that it is.
08Your choices
You decide which sources to connect, and you can disconnect any of them. Working profiles are optional. You can ask for a copy of the information we hold about you, ask us to correct it, or ask us to delete it.
Depending on where you live, you may have further rights under local privacy law. Contact us and we will respond.
09Changes to this policy
If we change this policy we will update the date at the top of the page. If a change materially affects how connected content is used, we will tell the people using the pilot before it takes effect.
10Contact
Questions and requests about this policy or your information: braun@brand.co.