Privacy policy
What qloudpulse keeps about you, why, who else handles it, how long it is kept, and how to have it deleted.
In force from TODO: date
TODO This is a draft for legal review. Remove this notice once the text is approved.
In short
- No adsWe sell nothing about you, and share nothing for advertising.
- No trackingThis site sets no cookies and has no analytics. Its access log holds no address.
- 30 daysafter your company is deleted, its data is gone.
We keep what it takes to sign you in and to draw your company's reports, and we keep it in AWS's Europe (Ireland) region. A problem you report, and an error a page of the reports hits, also go to our engineers' private issue tracker on GitHub.
Who we are
qloudpulse is run by TODO: legal entity name, TODO: registered address. We decide what is collected about you and why, which makes us the controller of your personal data under the GDPR.
What your company keeps in qloudpulse, such as what we collect from its AWS accounts, is your company's. We handle it on your company's behalf, and its admins decide about it, under our data processing agreement.
For anything about your personal data, write to TODO: privacy contact email. For help with the product, write to support@qloudpulse.com.
What we collect, and why
When you sign up with your email
- Your first and last name, and your work email address, which is also your username.
- Your password. Keycloak, the sign-in service we run ourselves, keeps only a hash of it, so nobody can read it back, us included.
- One email, to check that the address is yours.
Why: to give you a login and let you sign in. It is needed to provide qloudpulse to you and your company.
The sign-up page also uses the address your browser connects from (your IP address) to allow no more than 10 sign-ups an hour from one address. It holds the address in memory for that hour and never writes it down.
Why: to slow down anyone creating logins in bulk, in our legitimate interest in keeping the service safe.
When you choose Continue with Google
- Nothing goes to Google before you click. Our sign-up page draws the button itself and loads nothing from Google.
- When you click, your browser goes to Google, and Google learns that you are signing in to qloudpulse. What Google does with that is covered by Google's privacy policy.
- We ask Google only for your basic profile and email address. Google sends your name, your email address, which Google has verified, and Google's id for your account. Keycloak keeps those with your login, so no email goes out to check the address.
- We keep no Google tokens, so qloudpulse cannot reach your Google account or read anything else from it. We take your name from Google once, at your first sign-in.
Why: so you can sign in with your Google account instead of a password. It is needed to provide qloudpulse to you.
Your company
- The company's name, who belongs to it, and each person's role: admin or member.
- Invitations: the email address an admin invites. Keycloak sends the invitation by email and keeps it until it is accepted or withdrawn.
Why: so that your company's people see its reports and nobody else does. It is needed to provide qloudpulse to your company.
Your company's AWS accounts
- When your company connects its AWS accounts: their ids and names, the read-only roles we use to read them, and the ExternalId that ties those roles to your company. If you connect your bill, where it is read.
- What the collector reads from those accounts, which the start page lists under What we read. It describes your infrastructure, not people, but resource names and tags can hold a person's name, so we treat it with the same care.
Until your company connects its AWS accounts, nothing is read from them.
Why: to draw your company's reports. It is needed to provide qloudpulse to your company.
When your company keeps its data in its own AWS account
- An admin can choose, under Where your data is kept, to keep what is collected for the company in an S3 bucket and an RDS database in one of its own AWS accounts, in a region it picks, encrypted with its own KMS key. We keep the account's id, the region, and how the last check and move went.
- We move the data across, and keep our copy for 7 days in case the move has to go back. Then we delete it. The other clouds' bills your admins upload are kept in that bucket too.
- What stays with us: your company, its members and settings, sign-in, what you connected, the collector's queue of runs and each run's log, and the audit trail.
- Our services in Ireland still read the data to collect it and draw your reports, as before, but no longer keep it.
Why: so your company decides where its cost data lives. The choice is your company's, and so is that account.
Feature requests
- The title, details and importance of a feature you ask for, and up to three images. We write each image again, which removes its metadata, such as where a photo was taken.
- Who asked, so the page can mark your own requests. Your colleagues see the request but not who sent it, and our staff see the company, not the person.
Why: to decide what to build next and to answer you, in our legitimate interest in building what our customers need.
We may publish a copy of a request on this site's Feature requests page, written by us, with nothing that names your company or anyone in it. The images, how much it matters to you and our answer to your company are never published.
When you ask for a cloud or service
- The cloud or service you ask for, such as Azure or Hetzner, and what you would use it for, if you say.
- Who asked, so the page can mark your own, and whether you want to hear when it is ready. We do not copy your email address for that: when it is ready we read it from your login at that moment and send you one mail. If you have left the company by then, you are not mailed.
The name you asked for, and how many companies asked for it, go to our engineers' issue tracker on GitHub, in the repository problem reports go to. Your company, you and what you would use it for do not.
Why: to decide which clouds and services to cover next and to tell you when yours is ready, in our legitimate interest in building what our customers need, and the mail because you asked for it.
When you star or comment on a feature request on this site
- A random token this page makes and keeps in your browser's storage, and the SHA-256 of it beside each request you starred, so a star counts once per browser. The token says nothing about you, and we cannot tell two browsers' tokens apart from one another in any other way.
- The text of a comment, as you wrote it, with the time. There is no field for a name or an address, and a comment is shown as written once one of us has read it, so please keep anything that identifies you out of it. A comment we do not show is deleted.
- Your IP address, counted in memory for an hour to limit how many stars and comments one address sends, and never stored or logged.
Why: to learn what people want built, in our legitimate interest in building what our customers need, and to keep a script from drowning out the people.
When you report a problem
- What you write: a summary, what happened, what you expected, and the steps that lead to it. It goes as you wrote it, so leave out what you would not send us.
- What the page you were on saw, which the report page shows you in full before you send it: the page and its title, the pages you opened before it, the release, the errors the page met and the calls to our services that failed, and your browser with its language, time zone, window size and light or dark mode. Email addresses, sign-in tokens and the values of your report's filters are left out of it, by your browser and again by us.
- Up to three screenshots. We draw each one again, which removes its metadata. A screenshot shows whatever was on your screen, so the page asks you to check it first.
- Your company's id. A report does not say who sent it: it holds no name, email address or login id. The audit trail records that you sent it, as below.
We keep the report and its screenshots in our own storage in AWS. Everything but the screenshots also becomes an issue in our engineers' issue tracker on GitHub, in a repository that only they can open. A company can send at most 20 reports a day.
Why: to find and fix what went wrong, in our legitimate interest in a product that works for the people who use it.
When a page hits an error
- What the error said and where in the page's code it happened, the page and its release, and your browser with its window size and light or dark mode. Email addresses, sign-in tokens and the values of the page's filters are left out of it, by your browser and again by us. It goes the moment it happens, without asking, since the point is to hear of the errors nobody reports.
- Your company's id, so we can look at the error with your company's data. It holds no name, email address or login id.
Each distinct error becomes one issue in our engineers' issue tracker on GitHub, in the same repository, and the same error again adds to that issue's count rather than a new issue. A company's pages can send at most 60 an hour.
Why: to find and fix what went wrong before anyone has to report it, in our legitimate interest in a product that works for the people who use it.
When you test qloudpulse for us
Some people test qloudpulse for us, and only they see its Test panel. Our staff choose who they are.
- What you write in the panel: what you were trying to do, what happened, a word you did not understand, how a task went and how easy it was. It goes as you wrote it, so leave out what you would not send us.
- The page you were on and its heading, the release, and your browser with its window size and light or dark mode. Email addresses, sign-in tokens and the values of your report's filters are left out of the page's address.
- Your company's id. It holds no name, email address or login id.
- Up to three images with a task, if you add them: a screenshot you paste or pick, or a picture of the page you capture. Capturing asks your browser to share the tab, takes one picture and stops sharing. An image shows whatever was on the screen, your company's figures included.
Each one becomes an issue in our engineers' issue tracker on GitHub, in the same repository as problem reports. The images do not: we keep them in our own storage in AWS, with problem reports' screenshots, where only our staff can open them. Which tasks you have finished stays in your browser. A tester can send at most 60 an hour.
Why: to learn where qloudpulse is hard to understand or use and to fix it, in our legitimate interest in a product that works for the people who use it.
When you write to support
- Your email address and name as your mail gives them, what you wrote and anything you attached.
- Mail to support@qloudpulse.com is received by Amazon SES, which keeps an encrypted copy in our AWS account for 30 days, and forwarded to our staff's support mailbox in Gmail, where the answer comes from.
Why: to answer you. It is needed to provide qloudpulse to you, or in our legitimate interest in answering people who write to us.
What our customers have saved
The front page shows what our customers' changes have saved, in total. It is worked out from each company's savings, added up across companies, and holds nothing about any person or any one company.
Why: to show what qloudpulse does for the companies that use it, in our legitimate interest in telling people about it.
The audit trail
- A record of every change to a company's users and settings, of each feature request and answer, of each problem report, of each deletion, and of our staff opening a company's pages, a problem report or a tester's images.
- Each entry says what happened and when, who did it (your login's id, or the name of a member of our staff), and whom it concerned, such as an invited email address.
- It is kept apart from the services, in storage that nobody can change, for up to 7 years.
Why: so that every change, and every look our staff take at a company's data, can be checked afterwards, in our legitimate interest in security. We are working towards SOC 2, which asks for this record.
In your browser
- This site, qloudpulse.com, sets no cookies, has no analytics and loads nothing from anywhere else, apart from two figures it reads from our own app at app.qloudpulse.com: the published feature requests, and the total saved on the front page. Those requests carry no cookie, and the app does not keep your address. Its access log, kept for 90 days, holds the page asked for, the time, the country the request came from and the browser's name, and never your address, the page's query or a cookie, so no line of it names you. We count page views from it, by day, page, country and the site a visit came from, and keep only the counts.
- Signing in to the reports uses cookies that Keycloak sets on auth.qloudpulse.com, which it needs to keep you signed in. While you are signed in, Keycloak keeps your session with the address you signed in from, and it counts failed sign-ins to stop anyone guessing your password.
- The reports keep your sign-in in the browser tab's session storage. A few of your choices, such as the AWS accounts you follow and the AWS CLI profile names in the commands they show, stay in your browser's local storage and are never sent to us.
Who else handles it
- Amazon Web Services runs qloudpulse, its databases and Keycloak, in its Europe (Ireland) region, eu-west-1. Our emails, such as address checks, invitations and password resets, go out through Amazon SES in the same region. AWS's CloudFront delivers this site from locations near you, and keeps no log of it for us.
- Google, if you choose Continue with Google, as described above, and if you write to support, since our support mailbox is in Gmail. TODO: Google's role and the safeguard for any transfer out of the EU, for legal review
- GitHub, only if you report a problem, a page of the reports hits an error, you send feedback from the Test panel as a tester, or you ask for a cloud or service, which sends only its name. Our engineers' issue tracker gets what you write and what the page saw, never the screenshots, in a repository that only they can open. GitHub is run by GitHub, Inc. in the United States, and keeps the issue there. TODO: GitHub's role (processor or controller) and the safeguard for the transfer to the US, for legal review
Nobody else. Your data is stored in the EU, in Ireland, unless your company chose to keep it in its own AWS account, where it is stored in the region your company picked, and apart from a problem report's, an error's or a tester's issue on GitHub, which is kept in the United States, and mail to support, which is kept in Gmail. A visit to this site may be answered from outside the EU by CloudFront, and a Google sign-in may be handled outside the EU by Google, under their own terms.
How long we keep it
- Your login
- Until an admin of your company removes you, which deletes it at once, or until your company is deleted. A login that belongs to no company, such as one made during early access before your company is opened, stays until you ask us to delete it.
- Your session
- Until you sign out, or it runs out.
- Invitations
- Until accepted or withdrawn, or until the company is deleted.
- Your company's data
- Until the company is deleted, and gone within 30 days after that, as below. Backups age out 35 days later.
- Feature requests
- As long as the company. If you leave it, your requests stay with the company, no longer linked to you.
- Clouds and services you asked for
- As long as the company. If you leave it, your ask stays with the company, no longer linked to you, and you will not be mailed.
- Stars and comments on this site
- As long as the request they are on is on this site, and the request's company exists; a comment we do not show is deleted when we read it. The token stays in your browser until you clear its storage.
- Problem reports
- As long as the company. They never say who sent them, so leaving the company changes nothing in them. When the company is purged, its reports and their screenshots are deleted, and their GitHub issues lose their title and text. GitHub keeps an issue's earlier text in its edit history, which our engineers can still see until the issue itself is deleted.
- Testers' feedback
- Until our engineers have acted on it and delete the issue, and at the latest when the tester asks us to delete it. It names no person, so leaving the company or a company's purge does not change it. Its images are your company's data: they are deleted when the company is purged, and sooner if the tester asks.
- Errors the pages hit
- An error's issue names your company's id and nothing about you, so it stays when the company is purged, as long as our engineers keep it. We keep a count of each error, without the company, beside the issue.
- Mail to support
- 30 days for the copy in our AWS account; in the support mailbox, as long as we need it to help you. TODO: a retention for the mailbox
- The sign-up limit
- An hour, in memory only.
- The audit trail
- Up to 7 years.
Deleting your company
An admin deletes the company on its page in qloudpulse, by typing its name again. Our staff can also delete it if you ask.
- At once: nobody can sign in to the company any more, within a minute, and the logins of everyone who joined it by invitation are deleted.
- Within 30 days: a purge that runs every day removes the company's database, with its reports and everything collected for it, the logins of its remaining users, its feature requests with their images, its problem reports with their screenshots, its AWS settings, and its name. Its problem reports' GitHub issues lose their title and text, with the edit history GitHub keeps, as above; issues about errors its pages hit stay, since they name only the company's id. What stays is a line saying that a company existed, with its id and the dates it was deleted and purged.
- Backups of what was deleted age out 35 days after the purge.
The audit trail keeps its record of the deletion, as above. The company's page then shows how to remove our read-only roles from your AWS accounts, which ends our access to them.
If your company keeps its data in its own AWS account, the purge ends our access but leaves the bucket and database there: they are your company's to keep or delete. Deleting the role qloudpulse-storage, or disabling the key, ends our access at any time.
How we keep it safe
- Everything is served over HTTPS only.
- Each company has a database of its own, and every request is checked against the company of the person signed in.
- We read your AWS accounts through a read-only role that you create, and that denies access to what is stored in them.
- If your company keeps its data in its own AWS account, we reach it through one role that trusts only your company, the database answers our addresses only and over TLS, and everything there is encrypted with your company's own key.
- Our staff's access to a company's data is recorded in the audit trail.
We are working towards SOC 2. We do not have a SOC 2 report yet.
Your rights
Under the GDPR you can ask us:
- For access to what we keep about you, and a copy of it.
- For correction of anything that is wrong.
- For deletion of what we keep about you.
You can also ask us to limit how we use your data, object to a use we base on our legitimate interest, and have the data you gave us in a common format.
Write to TODO: privacy contact email, from the address of your login if you can, so we can tell the request is yours. We answer within one month. Where it is your company's data, such as what we collect from its AWS accounts, we work with your company's admins, who can also remove people and delete the company themselves.
If you think we handle your data wrongly, you can complain to the data protection authority in the EU country where you live or work, or where you think it happened.
Changes to this policy
When this policy changes, this page changes with it, and so does the date at the top.