← Back to Senshin

Privacy Policy

Last updated: 10 November 2026 · Version 3.5 · Terms of Service

Two separate things. Your project (names, RAID text, files, exact pounds) stays in your workspace and is used to run Senshin, including Sensei. Separately, learning from programmes is off for every organisation until its owner or an administrator turns it on and agrees. If they do, we keep only bands, counts and outcome codes, never names or text. The data in a workspace belongs to the organisation that pays for it. We do not sell personal data. We do not train foundation models on your project text.

1. Who we are (controller)

The data controller is Extreme Gravity Ltd, trading as Senshin (senshin.io), a company registered in England and Wales, company number 09590911, registered office 28 Windermere Avenue, London N3 3QY.

Privacy: privacy@senshin.io

When you put other people's data into a project (colleagues, stakeholders, vendors), you are usually the controller of that data and we are your processor. That relationship is in the Data Processing Agreement.

The data in a workspace belongs to the organisation that pays for it. For learning from programmes (section 3) we are the controller of the coded numbers we keep for that purpose, and the organisation decides whether to allow it.

2. What we collect

Account data

Name, email, organisation, role and timezone, so we can create the account and bill it. Sign-in is by an emailed code, Google or Apple, as offered on the sign-in screen.

At sign-up we may derive an approximate location (country and city) from the connection's IP address, then discard the address. We keep only the country and city, not the address.

Project delivery data

What you enter or import to run a project: names, dates, methodology, RAID, finance, people and RACI, activities and dependencies, files and artefacts, Briefs and SteerCo packs, and comments you leave in the product.

Sensei and product telemetry

On our own systems we keep a record that an analysis ran (who, which project, which depth, whether you accepted or dismissed a nudge). That is our audit trail, not a copy of the prompt at the model provider. Usage telemetry (pages, features, device class, timing) is first-party. Optional analytics cookies are off unless you consent — see the Cookie Policy.

Visit recordings

To improve the site, we record some visits (about half) using our analytics provider. Everything you type is hidden, and payment and sign-in fields are never recorded. Public pages set no cookies, and we respect Do Not Track. On public pages we also record how long a page was open and how far down it was scrolled, anonymously, without cookies, browser storage or any identifier. To ask us to delete a recording, email ops@senshin.io.

Billing

We do not store card numbers. Our payment provider processes payment. We store the customer reference, plan, seat count and invoice metadata we need to run the subscription.

3. How we use it and on what legal basis

PurposeDataLegal basis (UK GDPR)
Run the product and your accountAccount, project dataContract
Sensei analysis and the BriefProject data you have enteredContract
Billing, refunds, invoicesEmail, plan, payment referencesContract; legal obligation (tax)
Security and abuse preventionSession tokens, limited connection dataLegitimate interests
Service email (trial ending, receipts)Email, nameContract
Learning from programmes (off unless your organisation chooses it)Bands, counts and outcome codes, with a coded programme reference while we hold the working copy; never names or textLegitimate interests, and only after your organisation's recorded choice; you can object
Optional analytics cookiesUsage eventsConsent
Service planningApproximate location (country, city) at sign-upLegitimate interests

Learning from programmes

This is optional and off by default. Senshin works in full without it.

The organisation that pays decides, not an individual user. Only an owner or administrator of that organisation can turn it on, in Settings → Your data. Nothing is pre-selected. Each choice is recorded with the exact words shown, who chose and when, and the organisation can download that history.

There are five levels. Levels 0 to 2 can be chosen. Levels 3 and 4 show as "Not available yet", and nothing uses data at those levels.

At Level 1, for each project of an organisation that has allowed it, we keep a short record made only of fixed choices and bands: type of programme, sector, method, phase, contract type, broad region, size bands for budget and team, how long it was planned for, how far it slipped or overspent in bands, counts of changes and of flags you accepted or dismissed, and how it finished. These come from four kinds of data: plan and schedule, risks and issues, budget and finance, and outcomes and close-outs. Sample projects are never used.

We never keep names, emails, free text, files, exact amounts or exact dates for learning. A record with a value that is not on a fixed list, or that looks like a name or a long number, is dropped.

While we hold it, the working copy carries a coded programme reference that only we can make, so we can delete it if your organisation stops. We treat it as personal data. It is kept apart from your workspace and deleted after at most 90 days, or as soon as it is mixed into an anonymous group, whichever comes first.

A group becomes anonymous statistics, and is shown to others, only when it holds at least 5 similar programmes from at least 3 different organisations. Until then, the benchmarks you see are labelled as sample benchmarks. Once a year we test whether any group could be traced back to an organisation, and we fix what we find.

The organisation can change or stop at any time in Settings → Your data. New contributions stop at once and we delete the working copy straight away; a nightly check deletes anything left behind. Numbers already mixed into an anonymous group cannot be taken back out.

While our learning terms are not in force, nothing is kept for anyone, whatever has been chosen.

We do not keep special category data (such as health, beliefs or trade union membership) for learning, and none of the fixed lists can hold it. We do not sell personal data or learning data. We do not show one organisation's records to another. Using customer data to train our own models would need an organisation to choose Level 3 or above together with a training purpose, and those levels are not available yet.

You can ask us to stop using data about you for learning, or to delete it, at privacy@senshin.io. If you are a person in someone's programme rather than the account holder, you can write to us directly.

4. How Sensei processes data

  1. Deterministic checks run on our infrastructure (schedule, budget, RAID, people, governance). That stage does not send your data to a model provider.
  2. New project files. We first read the file ourselves (cells, Gantt, tables). Running prose that still needs a language model is sent to an AI processing provider.
  3. Chat and narrative. To answer you in Sensei chat, or to turn live signals into readable analysis, we send the relevant project context to an AI processing provider under contract. Every model request goes through one AI gateway (direct calls to model companies are blocked in code), and every request tells it to route only to model hosts that do not store or train on prompts.

We do not use your project text, names or files to train foundation models. We cannot control a provider if they break their contract; we choose providers and settings to make that a breach of theirs, not our default.

Each paid seat includes £2 of AI use per month. Past that, Sensei uses the built-in deductions instead of a paid model. See Fair Use and the Terms.

5. How we keep it safe

Your data is held with trusted hosting and storage providers. It is encrypted in transit (TLS) and at rest, and access is controlled and limited to your workspace.

6. Who we share with

We do not sell personal data. We do not use advertising networks. Anonymous statistics are not personal data once a group is mixed; we do not sell your project.

We use trusted service providers (processors) to run Senshin, for hosting and storage, content delivery and bot protection, sign-in, payments, email and text messages, AI processing, error monitoring, and product analytics. They handle personal data only on our instructions and under written terms. We will tell you which providers we use if you ask: privacy@senshin.io. The processor terms are in the DPA.

7. How long we keep it

8. Your rights

Under UK GDPR you can ask to access, correct, erase, restrict, port, or object, including to the use of data about you for learning from programmes. Email privacy@senshin.io. We respond within 30 days.

You can also complain to the ICO.

In-product: Settings for export and account delete; Settings → Your data for learning from programmes.

9. Cookies

Necessary cookies run the session. Optional analytics and marketing cookies are off until you consent. Full list: Cookie Policy.

Until you choose, and if you choose “No thanks”, a public page is counted without a cookie, without browser storage and without an identifier. If you choose “Accept”, our analytics provider stores an anonymous ID in a first-party cookie and in browser storage, so visits to different pages can be linked to one browser. If you then sign up or sign in, that ID is linked to your internal account number (not your email address) so we can see which pages led to a trial. You can change your choice at any time with “Cookie settings” at the bottom of any public page; choosing “No thanks” later removes the analytics cookie and storage from that browser.

10. Mobile app

If you install the iOS or Android wrapper: a local encrypted copy can sit on the device for offline use; we do not back that copy up to iCloud or Google Drive. Push notification tokens are stored against the account so we can send alerts you have asked for. Biometrics stay on the device. Deleting the app does not delete the server account — use Settings → Delete my account for that.

11. Children

Senshin is a professional tool. It is not for anyone under 18. We do not knowingly collect children's data.

12. International transfers

Some of our providers, including AI model hosts, may handle data outside the UK. Where they do, we rely on an adequacy decision, the UK International Data Transfer Agreement, or the UK Addendum to the EU Standard Contractual Clauses (and the EU Standard Contractual Clauses for EU data).

13. Security

Session cookies are httpOnly and Secure where the site is served over HTTPS. Access inside the product is limited to your workspace. Report vulnerabilities to security@senshin.io. We do not claim a named certification (for example SOC 2) that we do not hold.

14. Changes

Material changes are notified in the product and by email with at least 14 days' notice. The date at the top is the current version.

15. Contact

Extreme Gravity Ltd, trading as Senshin
Company number 09590911 · Registered office: 28 Windermere Avenue, London N3 3QY
privacy@senshin.io · security@senshin.io · support@senshin.io