Privacy Policy

Draft · Version 0.1 · Not yet in effect

This policy explains what personal data Iter8 handles, why, who else touches it, how long it is kept, and the rights you have. We have tried to write it in plain English. If anything is unclear, email us and we will explain.

Who we are

Iter8 is provided by Shattered Sun Ltd, a company registered in England & Wales under company number 08517320.

  • Registered address: Belmont Suite, Paragon Business Park, Chorley New Road, Horwich, Bolton, England, BL6 6HG
  • Contact: hello@iter8.co.uk
  • Website: iter8.co.uk
  • ICO registration number: [number]

Two roles: ours and your organisation's

Iter8 is used by organisations. Data protection law gives us a different role depending on whose data it is and why we hold it.

  • We are the controller for the data we need to run your account and our business. That means your name, email address and sign-in, your organisation's membership records, our emails to you, our website, support conversations, billing, and the technical records that keep the service running. This policy is the full account of how we use that data.
  • We are a processor for everything your organisation puts into Iter8, or asks Iter8 to produce. That includes meetings, their recordings and transcripts, tickets, comments, documents, code, and the summaries and drafts our AI features write. We call this customer content. Your organisation is its controller and decides what goes in, who can see it and how long it stays. We use it only to provide the service to that organisation, under our Terms and Conditions and Data Processing Addendum.

If your question is about customer content, such as a meeting you took part in or a ticket that mentions you, please contact the organisation that runs that workspace first. We will help them answer you. If you cannot reach them, or you are not sure who they are, write to us.

What we hold about you

When you have an account

  • Your name and email address, which you give us when you sign up or are invited.
  • Your sign-in. We use Clerk to sign you in, with a one-time code sent to your email or through an identity provider you choose. We hold the identifier Clerk gives us. We never see or store a password, because there is none.
  • Your organisations and projects: which ones you belong to, your role in each, and the invitations you have sent or received.
  • Your preferences, such as whether to join meetings with your camera or microphone on.
  • What you do in the product, as part of the record of each change. Every item you create or edit carries who made the change and when.

Your avatar is drawn from your initials on our own servers. It is not fetched from any outside service.

When you take part in a meeting

Iter8 hosts video meetings and records them so they can be transcribed and turned into notes and tickets. Recording happens on Iter8's own meeting room, and the screen you join from says so before you join. For each meeting we hold:

  • Who was invited: name, email address, whether they accepted or declined, and whether they were invited in advance or added during the call.
  • When each person joined and left.
  • An audio recording of each person's microphone, kept as a separate file per person and deleted as soon as it has been transcribed. We do not record video or screen shares.
  • A transcript of what each person said, labelled with the name on their invitation.
  • What our AI features produce from the transcript: a cleaned-up account of what was said, the topics discussed, a summary, actions and draft tickets.

Speakers are identified by which microphone the sound came from, not by analysing anyone's voice. We do not create voiceprints, and we do not use voice to identify or authenticate anyone.

All of this is customer content. The organisation that ran the meeting is its controller and is responsible for telling participants that the meeting is recorded and for having a lawful reason to record it. Our Terms require this of them.

When your organisation connects its tools

An organisation can connect Iter8 to GitHub, Azure DevOps or Bitbucket by giving us an access token. With that access we read, and hold as customer content:

  • Repository details: names, branches and the latest commits.
  • Tickets imported from a tracker, including their text, status and the name of the person assigned.
  • For release notes: commits, pull requests, reviews and review comments in the chosen range, with the names and usernames of the people who wrote them.
  • For builds: a copy of the repository's code, described under "The coding agent" below.

Some of the people named in this data, such as a commit author or a ticket assignee, may not use Iter8 at all. Their names reach us because the organisation connected a tool that contains them. The organisation is the controller of that data.

Other things your organisation puts in

These are all customer content:

  • tickets and their comments
  • actions
  • documents and images uploaded to the asset library
  • the project wiki
  • the organisation's logo

How our AI features use your data

Several features send customer content to an AI provider to do their work:

  • Speech-to-text. Each meeting recording is sent to Deepgram, Inc. in the United States, which returns the transcript.
  • Meeting notes and tickets. The transcript is sent to Anthropic PBC in several steps to produce a summary, actions and draft tickets.
  • Release notes. Commits, pull requests and linked tickets are sent to Anthropic to describe what changed.
  • Wiki. Documents in a project's wiki are sent to Anthropic to build and update its pages.
  • The coding agent. Described in the next section.

Our commitments:

  • No training. We do not use customer content to train AI models. Our AI and speech-to-text providers are not permitted to train their models on it under their commercial terms with us. We have set both up to keep it for as short a time as they offer. Deepgram does not store the audio once it has returned the transcript, and Anthropic keeps what we send for [N days], [or not at all under a zero-retention agreement].
  • Drafts, not decisions. Tickets written from a meeting are drafts. They stay with the meeting until someone in your organisation publishes them. Nothing our AI writes makes a decision about any person.
  • AI can be wrong. Summaries, tickets and release notes can be inaccurate or incomplete. If something about you is wrong, it can be corrected or deleted like any other content.

The coding agent

An organisation can ask Iter8 to work on tickets in its connected repositories. When it does:

  • We copy the code. We take a copy of each linked repository's default branch onto our servers.
  • We prepare the agent's brief. We write a snapshot of the requested tickets for the agent to read, including the latest comments on each.
  • The agent works in isolation. It is an AI coding agent run on Anthropic's models, in a sealed container on our servers. It can read and change the copy of the code and reach the internet. It cannot reach the rest of our systems, and it never receives your organisation's access tokens.
  • We publish only to a new branch. When the agent finishes, we commit its changes to a new branch named for that build and open a draft pull request. We never push to an existing branch, and nothing is merged without a person in your organisation choosing to merge it.
  • We report back. The outcome is posted as a comment on each ticket.

The copied code and the agent's working files are deleted as soon as the build finishes. Only the outcome, the pull request link and the comments stay, as customer content.

Keeping the service running

To find and fix faults, our servers keep technical records of what they do: which kinds of request were made, how long they took, whether they failed, and background job progress. These records exist to tell us whether the service is working. We do not use them to profile anyone.

  • What they can contain: internal identifiers for accounts and items, and error messages, which can occasionally quote a few words of the content involved.
  • What they do not contain: IP addresses and browser details. We strip both before the records leave our server.
  • Where they are kept: Honeycomb and Grafana Labs hold them for us. Both are listed below.
  • How long: they are deleted automatically after no more than 60 days.

Cookies and storage in your browser

We use only what the service needs to work:

  • Sign-in cookies set by Clerk, which keep you signed in.
  • Your browser's local and session storage, which remembers which camera and microphone you picked and whether you are in a meeting, so a page refresh does not drop you.

We use no advertising or analytics cookies and no tracking pixels. Our web fonts are served from our own servers. Because everything here is strictly necessary, there is no cookie banner.

The providers we rely on

ProviderWhat they do for usData they receiveWhere
Clerk, Inc.Signing you inName, email, sign-in eventsUnited States
LiveKit, Inc.Running live meeting audio and video, which it relays and does not storeLive audio and video, display namesThe LiveKit server nearest each participant, which may be outside the EU
Cloudflare, Inc. (R2)Holding each recording until it is transcribedAudio recordingsEuropean Union
Deepgram, Inc.Turning recordings into transcriptsAudio recordingsUnited States
Anthropic PBCAI features and the coding agentTranscripts, ticket text, code, documentsUnited States
Resend, Inc.Sending invitation emailsThe invitee's email address and the organisation's nameUnited States
Honeycomb (Hound Technology, Inc.)Holding technical recordsAs described aboveEuropean Union
Grafana Labs (Raintank, Inc.)Holding the same technical recordsAs described above[EU region]
OVH SASOur servers and databaseEverything held in Iter8European Union

GitHub, Azure DevOps and Bitbucket are not on this list because they are not our providers. Your organisation connected them, and data moves between them and Iter8 on its instructions. The same applies to a tracker your organisation imports tickets from.

We will give organisations at least 30 days' notice before adding or replacing a provider that handles customer content.

How we protect your data

  • In transit: all traffic between your browser and Iter8, and between Iter8 and its providers, is encrypted with TLS.
  • Passwordless sign-in: there is no password to steal.
  • Access checks: every request is checked against your organisation and your project before any data is returned. Someone outside a project cannot tell whether it exists.
  • Our database: it is not reachable from the internet. Only our application can connect to it.
  • Recordings: they are encrypted at rest by our storage provider, and any read of one must be signed by our server.
  • The coding agent: it runs in a sealed container with no route to our other systems and none of your tokens.
  • Staff access: administrative access to production is limited to the people who run the service. We look at customer content only when your organisation asks us to, to investigate a fault, or when the law requires it.

No online service can promise perfect security. If a breach puts your rights at risk, we will tell the Information Commissioner's Office and the organisations affected, and you where the law requires it.

Why we use your data: our legal bases

This covers the data we control. For customer content, the organisation that controls it decides the legal basis.

What we doLegal basis
Creating and running your account, signing you in, sending invitationsContract, with you or with the organisation you belong to
Keeping the service secure, preventing abuse, fixing faultsLegitimate interests: running a safe, working service
Contacting you about your account or changes to the serviceLegitimate interests
Billing, once paid plans existContract, and legal obligation for tax records
Keeping records the law requiresLegal obligation

International transfers

Several of our providers are in the United States. Where a provider is certified under the EU-US Data Privacy Framework and its UK Extension, we rely on that. Where it is not, we rely on the Standard Contractual Clauses with the UK Addendum, or the UK International Data Transfer Agreement, in that provider's data processing agreement. You can ask us for a copy of the safeguards that apply.

How long we keep data

DataHow long
Your accountWhile your account exists, then deleted within 30 days of it being closed
Customer contentWhile the organisation exists, or until the organisation deletes it. Deleted within 30 days of the organisation closing
Meeting audio recordingsDeleted as soon as the recording has been transcribed. The transcript and notes are kept as customer content
Copies of code made for a buildDeleted as soon as the build finishes
Technical recordsNo more than 60 days
Billing recordsSix years, as UK tax law requires

Your rights

You have the right to:

  • Access: a copy of the data we hold about you.
  • Rectification: correction of data that is wrong.
  • Erasure: deletion of your data.
  • Restriction: limits on how we use it.
  • Objection: to object to use based on our legitimate interests.
  • Portability: your data in a portable format.

To use any of these for data we control, email hello@iter8.co.uk. For customer content, ask the organisation that controls it. We will pass on any request that reaches us and help them answer it.

Complaints

If you are unhappy with how we have handled your data, tell us first at hello@iter8.co.uk. We will acknowledge your complaint within 30 days and tell you what we are doing about it.

You can also complain to the UK's Information Commissioner's Office at ico.org.uk, or, if you are in the EU, to your local data protection authority.

Closing your account or organisation

[Until this can be done in the product, email hello@iter8.co.uk and we will do it for you.]

  • Closing your account removes you from every organisation and deletes your account. Content you created inside an organisation belongs to that organisation and stays with it, credited to your name as it was.
  • Closing an organisation deletes all of its customer content, including meetings, recordings, tickets, documents and code copies, within 30 days. Organisation owners can export their content before then.

Automated decisions

We make no decisions about anyone by automated means that have legal or similarly significant effects, and we do not profile anyone. Our AI features summarise and draft. People decide what to publish and what to merge.

Who Iter8 is for

Iter8 is a service for businesses and other organisations. It is not intended for anyone under 18.

Changes to this policy

This policy is versioned and dated at the top. If we make a material change, we will email the owners of every organisation and say so in the product at least 30 days before it takes effect.