Privacy Policy
In this document
Introduction
SorviAI is business software. We collect the information we need to give you an account, keep it secure, bill you and support you. We do not sell your personal information, and we do not use your business data to advertise to you.
Most of the information inside SorviAI is not ours. It belongs to the organisation that put it there. When your employer or a supplier uses SorviAI to store records about you, they decide what happens to that information and we hold it for them. The section on our two roles explains this properly, because it changes who you should contact about your data.
The SorviAI application, its database and the files uploaded to it are held in London. A small number of named suppliers help us run the service, and they are all listed under who we share information with. If you want your data corrected or deleted, your rights tells you how to ask, and which law applies explains the law behind them. If you are unhappy with how we have handled your information, complaints explains how to raise it with us and with the regulator.
Who we are
SorviAI is a business software platform operated by SORVI AI LTD, company number 16927720, registered in England and Wales, with its registered office at Unit 5a Victoria Road, Stoke-On-Trent, England, ST4 2HS. Our full company details are on our company information page.
In this policy, “we”, “us” and “our” mean that company. “You” means the person reading it, whether you hold a SorviAI account yourself or your details appear in someone else’s account.
We are registered with the Information Commission (the ICO) under registration number ZC254426. You can check that entry on the ICO’s public register of fee payers. Our data protection contact is [email protected], and how to contact us lists our regional contacts.
Our two roles, and why it matters to you
SorviAI is used by organisations to run their business. That means we handle personal information in two very different ways, and your rights depend on which one applies to you.
When we decide, we are the controller
If you sign up for SorviAI, get invited into a workspace, pay us, visit our website, subscribe to our newsletter, apply for a job with us or contact our support team, we decide why and how your information is used. In legal terms we are the data controller, and this policy is the document that governs it.
When our customer decides, we are the processor
If your details are inside SorviAI because an organisation put them there, that organisation is the controller and we are only the processor. This applies if you are an employee whose record sits in the Worklink module, a customer or supplier whose contact and invoice records sit in Finance, someone who reported a repair ticket, a visitor signed in at a business’s front desk, someone who sent an enquiry through a storefront, someone who signs in to a business’s customer or vendor portal, someone who opens a file a business shared with a public link, or someone who exchanged emails with a business using our TeamEmail module.
In those cases we hold your information on that organisation’s instructions. The written terms that govern this are our Data Processing Agreement, which UK GDPR article 28 requires and which forms part of our Terms of Service. We do not decide what is collected, we do not use it for our own purposes, and we do not change or delete it on our own initiative.
What this policy covers
This policy applies to the SorviAI platform and every application within it, together with our public website and our sales and support communications.
What it does not cover
- Online stores run by our customers. Businesses can publish their own storefront using SorviAI. Storefronts show the business’s own catalogue and take enquiries; they have no checkout. If you are browsing one of those sites or sending an enquiry through it, the business running it is the controller of your information and its own privacy policy applies, not this one.
- Third-party services our customers connect. If a customer links a Microsoft 365 mailbox, a payment gateway, a carrier account, HMRC or another external account to SorviAI, that provider’s own terms and privacy policy continue to apply to the data held on their systems.
- Other websites we link to. We are not responsible for how they handle your information.
Information we collect about you
This section covers information we hold as controller, which means information about you as a user or prospective user of SorviAI.
Account and profile information
When you create an account or accept an invitation to a workspace, we collect your email address, first and last name, and optionally your phone number and a profile picture. Your profile can also hold additional email addresses and phone numbers, your gender, country, state or region, time zone and language, if you or an administrator add them. While you are signing up, before your email address is verified, we hold a short-lived pending sign-up record. We record whether your email address has been verified and when, when your account was created and last updated, and which workspaces you belong to.
Authentication and security information
To keep your account secure we hold your password in hashed form, never as readable text. If you enable two-factor authentication we store the secret needed to verify your codes in encrypted form, your recovery codes, and the time you last used a code so that the same code cannot be replayed. We also issue short-lived tokens when you sign in, switch workspace or accept an invitation.
Device, network and activity information
When you take actions in SorviAI we record what was done, when, by which account, together with the IP address and browser user-agent string of the request. This forms the audit log, which lets workspace administrators see who changed what and lets us investigate security incidents.
Each time you sign in we also keep a session record with the IP address, your browser’s user-agent string, a location label derived from the IP address (such as the country or network), and when the session was last used. You can see your own sessions in your profile. A session record is deleted after the session expires.
When someone checks in or out in Worklink, the IP address of that check-in and check-out is stored with the attendance record. That record belongs to the employer, and is described under employee records below. Where a workspace administrator has turned on IP-based access restriction for check-in, we also compare the address of each attempt against the addresses that administrator has allowed.
Error monitoring
When something goes wrong in the SorviAI app, an error report goes to our error-monitoring provider, Sentry, which processes it in its EU region in Frankfurt, Germany. A report carries technical details of the request and the fault, and identifies the account involved by its user ID, email address and username, together with the name of the workspace. Some error log lines can also contain the email addresses a message was being sent to.
Sentry’s session replay feature records what happens on screen to help us find the cause of a fault. In the SorviAI app it records 10% of sessions and every session in which an error occurs. It masks what you type into form fields, but it does not mask the text shown on screen or images, so a recording can show the records you had open, such as customer names, invoice lines or employee details. It runs only if you allow analytics in our consent banner, and stops when you withdraw that consent. While it runs, Sentry writes an entry named sentryReplaySession to your browser’s session storage.
Billing information
We hold your billing contact details, the plan and subscription associated with your workspace, and a record of invoices and payments. Subscription payments are processed by Stripe. Card numbers go straight to Stripe and never reach our servers; to show you which card is on file we fetch its brand, last four digits and expiry date from Stripe when the page loads, and we do not store them.
Support and correspondence
If you contact us for help we keep the messages you send, our replies, and any diagnostic information you choose to share so that we can resolve the issue and improve the product. If you report a problem with the bug reporter inside the app, the report reaches our own support desk as a ticket, together with the details the app attaches: the screen you were on, app versions, your browser details, recent failed requests, the Sentry error reference and any screenshot you include. Our support mailbox runs on Microsoft 365.
Website information
When you visit our public website we collect standard technical information such as IP address, browser type and the pages you viewed, from our server logs. We run no advertising scripts on that site. If you allow analytics in our consent banner, Google Analytics also records the pages you view, the page that referred you, an approximate location derived from your IP address, your device and browser details, and how you engage with the site, such as scrolling, following links that leave it, searching it, downloading files and playing videos. Nothing from Google loads before you allow it. The script of Tradedoubler, our affiliate network, loads only if you allow affiliate tracking. Both are explained under who we share information with. Our network provider, Cloudflare, sets a short-lived security cookie to tell people from automated traffic; the other cookies the site can set are written only after you act: for example, your answer in the consent banner, your answer to a “was this helpful?” question on our articles and guides, or the country you pick in the country switcher. Our Cookie Policy lists each of them. When you arrive without having picked a country, the site may send you to a country version based on the country Cloudflare derives from your IP address; that country is not stored in a cookie or linked to you.
If you open the chat with our assistant, that conversation is kept in your own browser so a reload does not lose it: it is answered by a script running there, it never reaches us, and closing the chat erases it. The assistant does not ask for or store your email address or phone number. You can also write to a desk from inside that chat, and a message you send that way reaches us exactly as one sent from our contact page does.
A message sent through a contact form arrives with us by email, together with the name and address you give us. Any file you attach is held in private storage and a copy is included in the email to the desk. Attachments are deleted from our storage automatically 30 days after they are uploaded, and the copy in the receiving mailbox is kept as set out under how long we keep information.
Newsletter
If you subscribe to our newsletter we hold your email address, your name if you give it, where you signed up, and a record of your consent: the time you asked to subscribe, the IP address the request came from, and the time you confirmed. We send you a confirmation email first and add you to the list only after you click the link in it, so that nobody can subscribe you without your knowledge. We record whether each issue was delivered to you. Newsletters carry no open tracking. Links in an issue pass through our own short redirect, which counts how often each link is followed without recording who followed it. Every newsletter carries an unsubscribe link, and unsubscribing takes effect at once.
Marketing to recipients in the United Kingdom
Under the Privacy and Electronic Communications Regulations 2003 (PECR, regulation 22) we may send marketing email to an individual, including a sole trader or a partnership in England, Wales or Northern Ireland that is not a limited liability partnership, only with prior consent or, for our own customers, where we collected the address in a sale or negotiation, the email is about our own similar products, and we offered a way to opt out when we collected it and in every message. Subscribers who sign up on our website confirm by email before they receive anything. Our earlier launch “notify me” list was single opt-in, and addresses imported from existing lists can be added as confirmed without a confirmation email; we send no marketing to those addresses until they have re-confirmed by email. Every message says who it is from and gives a working address for opting out (regulation 23).
Job applications
If you apply for a role through our careers page we hold the name, email address and phone number you give us, the role you applied for, your CV, and a cover letter if you choose to send one. We also record the time you gave your consent and the IP address you gave it from, so that we can show the consent was given. We ask for your consent before you send any of it, and we rely on that consent as our legal basis.
Your CV and any cover letter are stored privately, are never published, and are reached only through a link that expires, by the people involved in hiring for that role. The hiring team is told about a new application by email; that email carries your contact details and time-limited links to your files, not the files themselves.
We keep an application for 12 months from the day it arrives and then delete it, together with the files. You can withdraw your consent or ask us to delete an application at any time, and we will act on it.
Information we process for our customers
This section covers content that our customers put into their workspace. We process it on their instructions, as described under our two roles. We are telling you what it consists of so that you know what we hold, not because we decide what happens to it.
Business contacts and financial records
Names, job titles, email addresses, phone numbers and postal addresses of a customer’s own clients and suppliers, together with quotes, sales orders, invoices, credit notes, bills, purchase orders, goods receipts, payments, expenses, inventory items, price lists, VAT returns and storefront enquiries. The same applies to budgets, statements, stock counts, point-of-sale sales (including walk-in customers), helpdesk contacts, companies and tickets in Escalate, call and meeting records, campaign contacts and repair technicians.
Employee records
Customers using the Worklink module can store detailed records about their staff. Depending on what that employer chooses to record, this may include a profile photograph, date of birth, gender, marital status, a national identification number, a tax identification number and a social security number (each a free-text field the employer may fill or leave empty; in the United Kingdom a National Insurance number will usually end up in one of them), emergency contact name, phone number and relationship, bank name, routing code and bank account number, annual pay, present and permanent home addresses, personal email address and mobile number, dependants with their names, relationship and dates of birth, education and work history, an “about me” text, uploaded documents such as contracts or copies of right-to-work documents, performance reviews, training records, company assets issued to the employee, resignation or separation requests with exit-interview answers, and day-to-day attendance, shift, leave and timesheet records. Worklink has no dedicated right-to-work field or document type: such a copy is an ordinary uploaded document.
Every workspace that uses Worklink is set up with a “Sick Leave” leave type, and every leave request needs a reason typed as free text, so an employee or manager can end up writing details of an illness there. A leave request can also carry a supporting document, such as a fit note, if the employee or employer attaches one. Details of an illness are health data.
There is no dedicated National Insurance number field: an NI number goes into one of the free-text identification, tax or social security fields. Those fields are stored, and returned by our programming interface, in plain text to anyone the employer has given access to employee records: there is no separate restriction on them. Detail screens show only the last four characters; the edit form shows the full number.
For every working day, a Worklink attendance record holds:
- the check-in and check-out times, and the hours, breaks and overtime calculated from them;
- whether the employee arrived late or left early, by how many minutes, and any reason the employee gave for being late;
- how the check-in was made (in the web app, from the header, by an administrator, automatically, through a personal link or on a wall-mounted kiosk tablet using the employee’s personal kiosk code), and which kiosk tablet, if one was used;
- the IP address of the check-in and of the check-out.
Worklink also shows each employee’s live presence status (whether they are checked in and when they were last active) to users with access, and it marks employees as absent automatically each day when no check-in is recorded, for the employer to review. The personal check-in link works without signing in.
Worklink uses no biometrics: no fingerprints and no facial recognition. It does not record location coordinates or photographs at check-in; the only image it holds is an optional profile picture. The IP address is stored on every check-in and check-out; there is currently no setting to turn that off.
Employers in the United Kingdom
You are the controller of your employees’ records. National Insurance numbers are needed for payroll only; copies of right-to-work documents are kept for the employment and for two years after it ends; working-time records are kept for two years under the Working Time Regulations 1998; and health information such as a fit note needs a condition under Schedule 1 to the Data Protection Act 2018 and, usually, an appropriate policy document. Our GDPR Compliance page sets out what that means for Worklink.
Visitors at a business’s front desk
A business using Worklink’s visitor desk can record each visitor’s name, phone number, email address, company, purpose of visit, the person they are visiting, badge, notes, check-in and check-out times and the office they visited. It can also record the type and number of an identity document the visitor shows, such as a passport or residence permit number. The document number is masked in responses from our programming interface. The business is the controller of these records.
Mailbox content
Where a customer connects a Microsoft 365 mailbox to the TeamEmail module, we sync and store message content, subject lines, sender and recipient addresses, attachments and threading metadata so that the customer’s team can work on those conversations inside SorviAI. Access is granted by the customer’s own administrator through Microsoft’s consent process and can be revoked by them at any time, which stops any further sync.
Emails a workspace sends
When a workspace sends a document by email from SorviAI, such as an invoice or a quote, we keep a send log: the sender, the recipients, the subject, the full rendered message, the time it was sent and whether it was delivered. Document emails are sent through Twilio SendGrid with open and click tracking switched on by default: each carries a tracking pixel, and its links are rewritten to pass through SendGrid. SendGrid records on its own systems when a message is opened and when a link is clicked. Our send log has fields for open and click times, but nothing fills them today, so those results do not reach SorviAI. Customer portal emails and workspace invitations are sent without tracking.
Service and repair tickets
Details of reported faults, the device or item concerned, and whatever the person reporting the issue chose to write, along with the contact details needed to keep them updated. The business’s staff can add comments to a ticket; those comments are internal to the business, and the person who reported the fault does not see the ticket or its comments in SorviAI.
Files and attachments
Any document, image or other file that a user uploads to their workspace, including files stored in Nest.
Files shared through a public link
A business using Nest can share a file with someone outside the business by creating a public link. Public links are allowed unless a workspace administrator switches them off. The person creating a link can protect it with a password, set it to expire, make it view-only, and revoke it at any time.
If you open a file through one of these links, we record each view, each download and each failed password attempt, with the time, the IP address the request came from and your browser’s user-agent string. The purpose is to let the business see how its link has been used and to detect abuse of the link, such as someone guessing its password. The business that created the link is the controller of this record and can see it; we keep it as its processor, until the link or the file is deleted. Revoking a link does not delete it.
How we use information, and our legal basis
Data protection law requires us to have a valid reason for each purpose. The table below sets out ours, in the terms used by UK and European law.
| What we do | Information used | Legal basis |
|---|---|---|
| Provide the platform and the applications you have subscribed to | Account, authentication, workspace content | Performance of our contract with you or your organisation |
| Authenticate you, prevent unauthorised access and investigate abuse | Credentials, IP address, user agent, audit records | Our legitimate interest in keeping the platform and our customers’ data secure. UK GDPR, as amended, names the security of network and information systems as an example of a legitimate interest; we still carry out the balancing test |
| Send service and security notices, such as downtime, breach or policy changes | Account contact details | Performance of our contract, and legal obligation where a notification is required by law |
| Bill you and collect payment | Billing and subscription details | Performance of our contract, and legal obligation for tax and accounting records |
| Answer support requests | Correspondence and diagnostic information | Performance of our contract, and our legitimate interest in supporting our product |
| Diagnose faults and improve the product | Error reports and aggregated usage information | Our legitimate interest in maintaining a reliable service |
| Record sessions to find the cause of errors (Sentry Session Replay), on 10% of sessions and every session with an error | A screen recording of your session, with form inputs masked but on-screen text and images not masked | Our legitimate interest in finding and fixing faults. The storage it writes in your browser needs your consent under PECR regulation 6, so it runs only after you allow analytics in the consent banner |
| Measure how many people visit our website and which pages they read (Google Analytics, on the website only) | Pages viewed, the referring page, approximate location derived from your IP address, device and browser details, and session and engagement events, linked to a random ID in the _ga cookie | Your consent under UK GDPR Article 6(1)(a), given in the consent banner, which PECR regulation 6 also requires for the cookies. You can withdraw it at any time through “Cookie settings” |
| Credit the partner whose link brought you to us, so they can be paid commission (affiliate tracking through Tradedoubler) | The Tradedoubler click reference from the partner’s link, kept with the workspace you create; for its first paid subscription, the subscription number, the first payment amount excluding VAT, and its currency | Your consent, given in the consent banner, which PECR regulation 6 also requires for the cookie that carries the click reference. You can withdraw it at any time through “Cookie settings” |
| Keep a record of each answer you give in the cookie consent banner, so we can show that consent was given or refused | A random consent ID, the optional categories you allowed, the button you used, whether you answered on the website or in the app, the page, the banner version and the time. No name, account, IP address or browser details | Legal obligation, because UK GDPR requires us to be able to demonstrate consent, and our legitimate interest in establishing or defending legal claims |
| Send marketing about our products, including our newsletter | Contact details, your preferences and the record of your consent | Your consent, confirmed by email, which you can withdraw at any time using the unsubscribe link in every message |
| Consider a job application | What you send us, and the record of your consent | Your consent |
| Answer a request from the police or another public body that needs information to prevent or detect crime, to safeguard someone, or to respond to an emergency | Only what the request needs | A recognised legitimate interest under UK GDPR, as amended by the Data (Use and Access) Act 2025, where the law allows us to rely on one; otherwise legal obligation |
| Meet legal and regulatory obligations, and establish or defend legal claims | Whatever is relevant and necessary | Legal obligation, and our legitimate interest in defending our rights |
Legitimate interests. Where we rely on legitimate interests, we have weighed our interest against your rights and freedoms and concluded that our use does not override them. You can ask us for a summary of that assessment at any time, and you can object to processing based on legitimate interests as described under your rights.
Recognised legitimate interests. Since the Data (Use and Access) Act 2025, UK law lists a small number of “recognised legitimate interests”, such as helping a public body with crime prevention, safeguarding or an emergency, for which no balancing test is needed. We rely on one only in the row above, and only for the disclosure the request needs. Marketing, internal administration and network security are named in the law as examples of processing that may be a legitimate interest, but they are not recognised legitimate interests: for those we still weigh your interests against ours.
Using information for a new purpose. We use information only for the purpose we collected it for, or for a new purpose the law treats as compatible with it. UK law, as amended in 2025, now lists some compatible purposes, and we will tell you before we use your information for anything else.
We do not sell personal information, we do not share it for cross-context behavioural advertising, and we do not disclose it to third parties for their own marketing.
Cookies and similar technologies
We use a small number of cookies and similar browser storage to make SorviAI work, and two optional categories, affiliate and analytics, that run only with your consent. A consent banner asks on your first visit to our website or the SorviAI app, and “Cookie settings”, in the website footer and the app’s user menu, lets you change your answer at any time. There are four categories.
- Strictly necessary. These keep you signed in, maintain your session, route you to the correct workspace and protect against cross-site request forgery. Our network provider, Cloudflare, also sets a cookie named __cf_bm on our website and on workspace addresses to tell people from automated traffic; it expires after 30 minutes of inactivity, and a cookie named sorvi_consent records your answer in the consent banner, with a random consent ID that links it to the record of that answer we keep on our servers. The service cannot function without these, so they are set without asking for consent.
- Preferences. These remember choices such as your light or dark theme setting and your layout preferences, and on our public website the country you picked and the conversation you opened with our assistant. They are written only after you act, and exist purely to make things behave the way you left them. PECR, as amended by the Data (Use and Access) Act 2025, expressly allows storage that remembers your chosen appearance or functionality without consent.
- Affiliate. If a partner’s link brought you to our website and you allow affiliate tracking, we keep the Tradedoubler click reference in a cookie named sorvi_tduid for 90 days, and Tradedoubler’s own script sets its tduid cookie and some local storage. Nothing is stored before you agree, and withdrawing deletes it. How the reference is used is explained under who we share information with.
- Analytics and diagnostics. Both run only if you allow analytics. On our public website, Google Analytics counts visits and the pages read, using the _ga and _ga_2L8QNHVNBH cookies, which last up to two years; switching analytics off deletes them. In the SorviAI app, Sentry Session Replay is a screen recording used to debug faults, described under error monitoring above. It masks form inputs but not on-screen text or images, and writes sentryReplaySession to session storage.
We show no adverts and set no advertising cookies, on the platform or on our public website. Google Analytics is used for audience measurement only, with its advertising features switched off. The 2025 amendments also allow some analytics without consent, if people are told and can object easily; we do not rely on that exemption. Google Analytics waits for your consent, and session replay is not simple statistics. Document emails that workspaces send are tracked by default, as described under emails a workspace sends; our newsletter carries no open tracking.
The SorviAI app currently loads its typefaces from Google Fonts, and loads the component that displays PDF previews, flag images, emoji images and the animation player from the jsDelivr content network, with unpkg as a fallback for the animation player. Opening the app, or the screen that needs one of these, sends your IP address and browser details to those providers. When you preview a Word, Excel or PowerPoint attachment, the app hands a signed, time-limited link to the file to the Google Docs Viewer (docs.google.com), which fetches the file to display it. None of these sets a cookie through our app.
Our Cookie Policy lists every cookie we set with its purpose and expiry, describes the browser storage we use, and explains how to clear or block it.
Who we share information with
We do not sell your information. We share it only with the suppliers listed below, who process it on our instructions, with our affiliate network if you allow affiliate tracking, with Google if you allow analytics on our website, and in the limited legal circumstances set out at the end of this section.
Our sub-processors
| Supplier | What they do for us | What they can reach | Location and transfer safeguard |
|---|---|---|---|
| DigitalOcean | Managed hosting, database, cache, background queue and object storage for uploaded files | All platform data, including attachments, documents, profile and item images | United Kingdom (London); database backups are kept off-site, at a location DigitalOcean does not publish. DigitalOcean, LLC is certified under the UK Extension to the EU-US Data Privacy Framework for non-HR data |
| Cloudflare | Network delivery and protection in front of our services, supplied through DigitalOcean’s App Platform | Traffic passing through, including IP addresses and request details | Cloudflare’s global network, answering from the data centre nearest the visitor. Cloudflare, Inc. is certified under the UK Extension to the EU-US Data Privacy Framework |
| Twilio SendGrid | Outbound and transactional email delivery | Recipient addresses and message content, and, for document emails, when each message is opened and each link clicked | United States. Twilio Inc. is certified under the UK Extension to the EU-US Data Privacy Framework for non-HR data; for HR data, the International Data Transfer Agreement in Twilio’s data protection addendum applies |
| Stripe | Payment processing for our own subscriptions | Billing contact details and payment card details, which go straight to Stripe | Stripe Payments Europe, Limited (Ireland), with Stripe Payments UK Ltd for regulated payment services in the UK: Ireland is covered by the UK’s adequacy regulations for the EEA. Stripe, LLC in the United States is certified under the UK Extension to the EU-US Data Privacy Framework |
| Functional Software (Sentry) | Error and performance monitoring | Technical error reports, which include the user ID, email address, username and workspace name and request details, and session replay recordings of on-screen content | EU region, Frankfurt, Germany: covered by the UK’s adequacy regulations for the EEA. Functional Software Inc. is also certified under the UK Extension to the EU-US Data Privacy Framework, for non-HR data |
| GeoNames | Finding the time zone for a postal code, used when arranging shipments | A postal code and country, with no name or account attached | Switzerland (Unxos GmbH): covered by the UK’s adequacy regulations for Switzerland |
| Google (Google Fonts) | Font delivery on the SorviAI app | The IP address and browser details of anyone opening the app | United States. Google LLC is certified under the UK Extension to the EU-US Data Privacy Framework |
| Google (Google Docs Viewer) | Displaying previews of Word, Excel and PowerPoint attachments | A signed, time-limited link to the file being previewed, and so the file itself, plus the viewer’s IP address and browser details | United States. Google LLC is certified under the UK Extension to the EU-US Data Privacy Framework |
| jsDelivr, with unpkg as a fallback | Delivering the PDF preview component, flag and emoji images and the animation player in the SorviAI app | The IP address and browser details of anyone loading those parts of the app | jsDelivr is operated by Volentio JSD Limited, a company registered in England and Wales, and served through global content networks (Cloudflare, Fastly and Gcore); unpkg is an open-source project served from Cloudflare’s global network. Requests are answered from the server nearest the visitor |
| Exchange-rate data provider (the free Currency API by default; ExchangeRate-API if selected in settings) | Fetching currency exchange rates, from our servers | Currency codes only; no personal data | Not applicable: no personal data is sent |
| Microsoft (Microsoft 365) | Our support mailbox | Messages you send to support and our replies | Microsoft’s data centres for our tenant. Microsoft Corporation is certified under the UK Extension to the EU-US Data Privacy Framework |
SorviAI does not currently use an AI provider; if one is added it will be listed as a sub-processor with its transfer basis.
Services a customer chooses to connect
These providers are not our sub-processors. A customer switches each one on for its own workspace, and the data they receive is sent at that customer’s direction and under that provider’s own terms.
| Service | What it is used for | What it receives |
|---|---|---|
| Microsoft 365 | Mailbox integration in TeamEmail | Mailbox content, at the customer’s direction, held in the customer’s own Microsoft region |
| Stripe (the customer’s own account) | Collecting payments from the customer’s own customers, when a payer pays an invoice online through the customer portal | Payment and payer details for the transactions that customer processes, including the payer’s email address and the invoice number. The payment is made to the customer’s own connected Stripe account, but the connected account is created, and each charge is made, through SorviAI’s own Stripe platform account (Stripe Connect) |
| HMRC Making Tax Digital (VAT) | Submitting UK VAT returns and viewing VAT obligations, liabilities and payments | VAT return figures and the device and connection details HMRC’s fraud prevention rules require with each call: the user’s public IP address and port with a timestamp, a hashed user ID, our own public IP address, a random device ID stored in the browser, the browser’s user agent, screen size, colour depth, display scaling factor and window size, and the time zone |
| UPS, FedEx and DHL | Creating shipments and labels | Sender and recipient names, addresses and contact details for each shipment |
Our affiliate network
Partners can earn commission for referring customers to SorviAI through Tradedoubler, our affiliate network. If you arrive through a partner’s link and allow affiliate tracking, Tradedoubler’s script on our website receives your IP address and browser details and sets the cookie and storage our Cookie Policy lists. When you create a workspace, we keep the click reference with it. When that workspace’s first paid subscription is completed, our servers send Tradedoubler the click reference, the subscription number as the order number, and the first payment amount, excluding VAT, with its currency, so the partner can be paid. We do not send your name, email address or payment card details, and later payments are not reported.
Withdrawing consent through “Cookie settings” deletes the cookies and storage in your browser. It does not reach a click reference already stored with a workspace; if you want that removed, write to us at [email protected].
Analytics on our website
We use Google Analytics to measure how many people visit our public website and which pages they read. It does not run in the SorviAI app. Its script loads from Google only after you allow analytics in the consent banner, and then sends Google the details listed under website information, linked to the random ID in its _ga cookie. Google states that Google Analytics does not log or store IP addresses. We use it for audience measurement alone: its advertising features are always switched off, Google Signals is not enabled, and we do no remarketing. Event-level data is kept for 14 months; the aggregated reports built from it are not subject to that limit. Google processes this data in the United States, and Google LLC is certified under the UK Extension to the EU-US Data Privacy Framework.
Withdrawing consent through “Cookie settings” tells Google Analytics that consent is withdrawn, disables it and deletes its cookies from your browser.
Changes to our suppliers
We review these lists when our suppliers change. If we add or replace a supplier that processes customer content, we will announce it on the Compliance page, which holds the maintained list, and email workspace account owners at least 30 days in advance, and a customer who objects may raise it with us under the Data Processing Agreement. There is no separate subscription for these notices.
Other disclosures
We may also disclose information where we are legally required to, for example in response to a valid court order or a lawful request from a regulator or law enforcement. In the United Kingdom such requests can come under the Investigatory Powers Act 2016, the Crime (Overseas Production Orders) Act 2019 or the UK-US Data Access Agreement. We disclose only what we are compelled to disclose, and where we are permitted to tell you or the customer concerned, we will. Our Abuse Policy describes how we handle these requests. If our business is sold or reorganised, information may transfer to the acquiring entity, which would remain bound by this policy or a materially equivalent one.
Where your data is stored, and transfers
The SorviAI application and its database run on DigitalOcean in its London region, behind Cloudflare’s network, and uploaded files are stored in DigitalOcean’s London region. Each customer’s data is held in a separate database schema, which is explained further under how we protect information. Every customer’s workspace is on the same infrastructure, unless we have agreed something different with them in writing.
Transfers out of the United Kingdom
Some of the suppliers named above operate outside the United Kingdom, and some of our staff provide support from India. Each kind of transfer rests on its own safeguard.
- Suppliers in the European Economic Area. Sentry processes our error reports in Frankfurt, Germany, and Stripe’s contracting entity is in Ireland. The UK recognises the EEA as adequate through its adequacy regulations, so no extra safeguard is needed. GeoNames is in Switzerland, which the UK also recognises as adequate.
- Suppliers in the United States. Each of our US suppliers is certified under the UK Extension to the EU-US Data Privacy Framework (the “data bridge”), and we rely on that certification for the data it covers. The same applies to Google Analytics data from our website, through Google LLC’s certification. DigitalOcean, Twilio and Sentry’s parent are certified for non-HR data only. For HR data, such as the Worklink records a customer keeps, Twilio’s data protection addendum applies the International Data Transfer Agreement, while DigitalOcean holds the data in London and its agreement adds no separate safeguard if that data is reached from the United States. The supplier table shows which applies to each.
- Access by our team in India. Some of our team work from India and can reach customer data to run and support the service. Everyone on that team is employed by SORVI AI LTD, so they are part of the same company, and the ICO does not treat their access to our UK systems as a restricted transfer. Before anyone works for us in India as a contractor or through another company, we will put the International Data Transfer Agreement in place with them, backed by a transfer risk assessment that covers access by Indian public authorities, because India has no UK adequacy regulations.
Since the Data (Use and Access) Act 2025, a transfer from the UK is lawful where the protection your information receives abroad is not materially lower than in the UK (the “data protection test”). The International Data Transfer Agreement and the UK Addendum to the EU Standard Contractual Clauses remain valid ways to meet it. For each transfer that rests on them, we will assess, before relying on it, that your information remains protected to that standard. You can ask for a copy of the safeguards that apply at [email protected].
How long we keep information
We keep information only as long as we need it for the purposes set out under how we use information, or as long as the law requires. Where nothing deletes something automatically, the row says so.
| Information | How long we keep it |
|---|---|
| Your account and workspace content, while the account is open | For as long as the account remains active. |
| Workspace content after a subscription ends | We do not delete a workspace automatically. When the customer asks us in writing, we delete it from our live systems within 30 days, and backup copies are overwritten within 7 days after that. A business may be required by law to keep its records for years after it stops using SorviAI, so it should export them first. |
| Contacts and customers deleted inside the product | A contact that has transactions cannot be deleted. When any other contact is deleted it is hidden and can be restored, and its name, email addresses, phone numbers, postal addresses and contact persons are kept; they are not cleared. Many other records, such as Worklink attendance records, are also hidden rather than removed when deleted. |
| Documents in Trash | Finance documents (such as quotes, invoices, sales orders and credit notes) and inventory adjustments go to Trash, where a workspace administrator can restore them. They stay in Trash until someone empties it, unless an administrator has switched on automatic deletion after 15, 30, 45, 60 or 90 days. Automatic deletion is off by default. Nest has its own trash, which is emptied after 30 days by default. |
| Backups | Seven days, after which backup copies are overwritten in the normal rotation. A backup is kept to recover from a failure. It is not an archive and does not replace any record a customer is required to keep. |
| Audit logs | For the life of the workspace. Nothing deletes them automatically. |
| Error monitoring reports and session replays | Up to 90 days, after which Sentry deletes them. |
| Affiliate click reference | In your browser, up to 90 days in the sorvi_tduid cookie (Tradedoubler’s own tduid cookie lasts up to a year), unless you withdraw consent sooner. On our servers, with the workspace’s account record, together with the order details once reported, until that record is deleted. Nothing deletes it sooner automatically. |
| Google Analytics data from our website | In your browser, up to two years in the _ga and _ga_2L8QNHVNBH cookies, unless you withdraw consent, which deletes them. At Google, event-level data is kept for 14 months; aggregated reports are not subject to that limit. |
| Email send logs, covering sender, recipients, subject, the full rendered message, timestamp and delivery outcome | For the life of the workspace, for delivery troubleshooting, audit and abuse investigation. Nothing deletes them automatically. |
| TeamEmail mailbox content synced from Microsoft 365 | Kept while the mailbox stays connected and the workspace is open. An optional purge that removes the body and attachments of archived messages after 90 days exists, but it is not switched on. |
| Sign-in session records (IP address, browser, location label) | Deleted after the session expires. |
| Worklink attendance records | Set by the employer, who is the controller. Worklink deletes none automatically. In Great Britain the Working Time Regulations 1998 require records adequate to show compliance to be kept for two years. |
| Record of views and downloads of files shared through a Nest public link (time, IP address, user agent) | Until the link or the file is deleted. |
| Sign-in and invitation tokens | Sign-in tokens expire within minutes of being issued. Invitations expire after seven days. Both are single use. |
| Invoices and accounting records about our own customers | Six years, as required by UK tax law. |
| Support correspondence | Six years from the last message. |
| Contact form messages | Attachments in our storage are deleted automatically 30 days after they are uploaded. The message in the receiving mailbox, which includes a copy of any attachment, is kept for six years. |
| Job applications | 12 months from the day the application arrives, then deleted with the files. If you join us, the application becomes part of your employee file. |
| Marketing and newsletter contacts | Until you unsubscribe, and then only a suppression record so that we do not contact you again. The record of your consent (time and IP address) is kept while you are subscribed and for six years after, so that we can prove it was given. Newsletter link counts are kept per link, not per subscriber, and are not deleted. |
| Data protection complaints | Six years from the date the complaint is closed. |
| Record of your answers in the cookie consent banner | Six years after your last answer, then deleted automatically, so that we can show what you agreed to for as long as a question about it could be raised. If you ask us to delete it sooner and quote the consent ID shown in “Cookie settings”, we will. |
Information we have deleted can remain in a backup copy until that copy is overwritten in the seven-day rotation. We do not separate it out in the meantime, and we do not use backups for anything other than recovering from a failure.
How we protect information
No system is perfectly secure, but these are the specific measures we take, and the limits we know about.
- Separation between customers. Every customer’s data lives in its own database schema, not in shared tables filtered by a column. Separation is enforced by the database itself, so a fault in application code cannot leak one customer’s records into another’s workspace.
- Encryption. Traffic between your browser and our servers is encrypted in transit. At rest, DigitalOcean encrypts the database volumes (LUKS, AES-XTS with 512-bit keys) and its database backups (AES-256), and states that files in its object storage are encrypted at rest.
- Private file access. Uploaded files are private by default and are served through signed links that expire, so a copied link does not grant lasting access. Files that a customer deliberately publishes, such as company logos and storefront product images, are served from public addresses that do not expire.
- Two-factor authentication. Every user can switch it on for their own account, with recovery codes. The secret used to verify your codes is stored encrypted, and each code can only be used once. Workspace administrators cannot yet require it of their users, it is not offered to Portal users, and hardware security keys and single sign-on are not supported.
- Short-lived sessions. Access tokens expire after 60 minutes and must be renewed.
- Role-based access control. Workspace administrators grant permissions per application, so a user only reaches the modules and records their role allows.
- Audit logging. Many actions are recorded with the account, timestamp, IP address and browser, so unusual activity can be traced. Coverage is incomplete: Repairs records only deletions, only a few settings screens write audit records, sign-ins through the public login, workspace selection and sign-in link exchange write no audit entry, and failed sign-ins are not logged with the account and IP address. The audit log is not yet tamper-evident, and an audit entry is not written in the same database transaction as the change it records, so a failed write can lose the entry.
- Staff access. When a member of our support staff opens a workspace as a customer’s user, which a senior engineer must approve and which is done only to resolve a support request, investigate a credible abuse report or meet a legal duty, the start of that session is recorded in the same transaction as the access, in our own platform audit log rather than the customer’s, and a banner shows in the customer’s interface. The individual actions taken during that session are not separately logged. Our director and one or two senior engineers can also connect to production databases directly; that list is reviewed every quarter, and those connections leave no audit record.
- If a breach happens. Where a personal data breach affects a customer’s data, we will tell that customer’s workspace administrators within 48 hours of confirming it, so that they can meet their own 72-hour deadline to notify the ICO. Where a breach affects information we control, we notify the ICO within 72 hours of becoming aware of it unless it is unlikely to result in a risk to you, and we tell you directly if it is likely to result in a high risk to you.
Our Security page describes each measure in more detail, and the same statements form the security annex of our Data Processing Agreement, so we list only what is true today.
Your rights
Under UK GDPR and the Data Protection Act 2018, you have the rights below.
- Access. Ask for a copy of the personal information we hold about you.
- Correction. Ask us to correct anything inaccurate or incomplete.
- Erasure. Ask us to delete your information, where we have no continuing lawful reason to keep it.
- Restriction. Ask us to pause our use of your information while a concern is resolved.
- Portability. Ask for the information you gave us in a structured, commonly used, machine-readable format.
- Objection. Object to processing we carry out on the basis of legitimate interests, and object at any time to direct marketing.
- Withdraw consent. Where we rely on consent, withdraw it at any time. This does not affect anything we did before you withdrew it.
- No penalty for asking. We will not deny you service, charge you a different price or give you a worse experience because you exercised a right.
To exercise any of these, write to [email protected]. We handle each request by hand: there is no self-service tool that gathers everything we hold about one person, so a member of our team searches each system and assembles the answer. We will respond within one month. If your request is complex we may extend this by up to two further months, and we will tell you within the first month if that happens. We do not charge a fee unless a request is manifestly unfounded or excessive.
We may need to verify your identity before acting, to make sure we are not disclosing your information to someone else. Under UK law, as amended in 2025, the one-month period starts once we have what we need to verify you; and if we reasonably need you to clarify what you are asking for, the clock pauses until you reply. When you ask for a copy of your information, we carry out a reasonable and proportionate search for it.
If your information is in a workspace run by your employer or another organisation, that organisation is the controller: send your request to them, and we will help them answer it.
Which law applies
This is the United Kingdom version of our Privacy Policy. It applies to workspaces set up for the United Kingdom and to visitors to our website in the United Kingdom. Everyone else is covered by the Global version.
| Where | What applies | What it adds | Who you can complain to |
|---|---|---|---|
| United Kingdom | UK GDPR and the Data Protection Act 2018, as amended by the Data (Use and Access) Act 2025, and PECR for marketing email and anything stored on your device | The rights listed above in full, with a one-month response deadline, and a right to complain to us about how we handle your information, which we must acknowledge within 30 days. | Us first, under complaints below; then the Information Commission (the ICO), ico.org.uk |
Across the United Kingdom
UK GDPR and PECR apply in the same way in England, Wales, Scotland and Northern Ireland. The Information Commissioner’s Office became the Information Commission on 30 September 2026 under the Data (Use and Access) Act 2025. It is still known as the ICO, and the complaint route is the same.
AI features and automated decisions
SorviAI does not currently use an AI provider; if one is added it will be listed as a sub-processor with its transfer basis.
We do not make decisions about you based solely on automated processing that produce legal effects or otherwise significantly affect you. One automated step exists: Worklink marks an employee as absent each day when no check-in is recorded. The employer reviews and can correct that record, and it is not a decision about the employee on its own.
Children
SorviAI is business software sold to organisations. It is not directed at children and we do not knowingly collect information from anyone under 16. We do not track anyone for advertising or target advertising at anyone, children included. In the United Kingdom the age of digital consent is 13, so our threshold of 16 is stricter than the law requires.
A storefront is run by our customer. If a storefront is likely to be used by children, the ICO’s Age Appropriate Design Code applies to the business running it.
If you believe a child has provided us with personal information, contact us at [email protected] and we will delete it.
Changes to this policy
We update this policy when the way we handle information changes. The effective date and version at the top always tell you which version you are reading.
If a change materially affects your rights or how we use your information, we will give workspace owners at least 30 days notice by email before the change takes effect. Minor corrections take effect when published. Version 1.3 adds Google Analytics on our website, which runs only if you allow analytics. It replaces version 1.2, effective 7 October 2026, which added the record we keep of your answers in the cookie consent banner and the automatic deletion of contact form attachments after 30 days. Version 1.1, effective 5 October 2026, added the consent banner and affiliate attribution through Tradedoubler and put Session Replay behind your consent. Every later version will be published here with its date, and earlier versions will be kept at this address.
How to contact us
For anything about this policy or your personal information, contact us using the details below. One address reaches us for everything. We acknowledge every privacy enquiry within 2 working days, and answer fully within the one-month deadline UK law sets.
Data protection contact
Regional contacts
Some of the regional laws set out above require a named contact inside the region. Where one is required of us, it is listed here.
| Role | Contact |
|---|---|
| Data protection officer | Not appointed. We are not required to appoint one: we are not a public authority, and our core activities do not consist of large-scale monitoring of people or large-scale processing of special category data. Deepak Gupta, director, is responsible for data protection. |
Complaints
If you are unhappy with how we have handled your personal information, please complain to us first, using the form on our contact page or by emailing [email protected]. We will acknowledge your complaint within 30 days, look into it, and respond without undue delay, telling you what we have done and what happens next.
You also have the right to complain to the regulator, the Information Commission (the ICO), at ico.org.uk or on 0303 123 1113. The ICO will normally expect you to have raised the matter with us first, but nothing here stops you going to it directly.
Key terms
The words in this policy that carry a specific legal meaning, in plain language.
Personal information
Any information that identifies you, or could identify you when combined with something else. A name and an email address obviously qualify. So do an IP address, an account identifier and a photograph. UK law calls it “personal data”; this policy says personal information, and means the same.
Controller and processor
Two roles defined by data protection law. The controller decides why and how personal information is used and carries the legal responsibility for it. The processor only handles it on the controller’s instructions and cannot use it for its own purposes.
SorviAI is the controller for your account with us, and the processor for the business records our customers keep in their own workspaces.
Legal basis
Under UK and European law, an organisation may not process personal information at all unless it can point to one of a fixed list of permitted reasons. The ones we rely on are performance of a contract, legal obligation, consent, legitimate interests and, in one narrow case, a recognised legitimate interest. The section on how we use information states which applies to each thing we do.
Legitimate interests
A legal basis that lets an organisation process information for a genuine business reason, but only after weighing that reason against your rights and deciding it does not override them. We use it for security, abuse investigation, keeping the service reliable and answering legal claims.
It is the most easily abused basis in the law, which is why we will send you a summary of the balancing exercise if you ask, and why you can object to any processing that rests on it.
Recognised legitimate interest
A UK-only basis added by the Data (Use and Access) Act 2025. The law lists a few purposes, such as helping a public body prevent crime, safeguarding a vulnerable person or responding to an emergency, for which an organisation does not have to carry out the balancing test. The list is short and fixed; ordinary business purposes are not on it.
Sub-processor
A supplier we bring in to help us deliver the service, which touches personal information in the process. The companies in our sub-processor table are ours; the services a customer connects are not. Each sub-processor must be bound by a written contract that places on it the same obligations we owe our customers, and we stay responsible to you for what they do.
International transfer
Moving personal information out of the country whose law protects it, or letting someone in another country access it. Under UK law a transfer is lawful where the destination is covered by UK adequacy regulations, or where safeguards such as the International Data Transfer Agreement or the UK Addendum keep the protection not materially lower than in the UK. Our US suppliers receiving data are transfers; our own employees reaching it from India are not, because they are part of our company. The section on where your data is stored names the footing for each.
Data Processing Agreement
The contract that has to be in place whenever one organisation processes personal information on another’s instructions (UK GDPR and GDPR article 28). It sets out what the processor may do, the security it must apply, what happens to the data at the end, and the terms on which sub-processors may be used. Ours is published and forms part of our Terms of Service.
Personal data breach
Not only a hacker getting in. It covers any security failure that leads to personal information being destroyed, lost, altered, disclosed or accessed without authorisation, including by accident and including by someone inside the organisation. Sending a file to the wrong customer is a breach. So is losing a backup.