Privacy policy
Last updated:
Controller
The controller for the processing of personal data within the meaning of the GDPR is Data Prudentia GmbH, Drötscher Straße 24, 56727 Mayen, Germany, represented by Wolfgang Hesse. Email: info@data-prudentia.de
For questions about the processing of your personal data and to exercise your data protection rights, please contact: Data Prudentia, email: datenschutz@data-prudentia.de.
We have currently not appointed a data protection officer.
This privacy notice covers the use of ucando by consumers. ucando is not directed at people under 18. If you use ucando through an account provided by your employer or another organisation, that organisation is the controller and we process the data on its behalf; in that case please contact them first.
Overview
ucando is an assistant you can use to write, ask questions, work through documents, generate images and search the web. Everything you type is sent to an AI model — that is the service, which is why the section "Processing by AI models" further down is the most important one here.
Four points up front, because they are the ones people ask about. First: we store your content in data centres inside the European Union; outside the EU, data is processed only by our payment provider Stripe, by web search (currently Microsoft Bing) and by Wikipedia when you have a term explained – details under "Transfers to third countries". Second: we do not use your content to train AI models. Third: we use no advertising or tracking technologies that recognise you — no ad networks and no analytics cookies, and for that reason no consent banner either. We count visits to our public websites without cookies (see "Audience measurement on our websites"). Our emails contain a counting pixel from our email provider that records opens and clicks as totals only, without linking them to you (see the "Service providers" page, sending emails). Fourth: the model provider may store suspicious inputs and responses to detect abuse; details are set out below.
What data we process
Account data. Your identifier, email address, display name, and your roles and permissions. We do not process your password: sign-in runs through our identity provider, which holds your credentials.
Content. Your messages, prompts and attachments, the models' responses, the structure of your conversations including branches, and titles and tags.
Documents. Files you upload, whatever they contain, the text extracted from them, its division into passages, and the vector representations used for search.
Projects. Your projects with their instructions and uploaded files and the conversations in them; every conversation in a project searches its files. A project is visible only to you. For each project you can set whether your personal memory is included; when it is off, nothing from memory flows into these conversations and nothing from them flows into memory.
Memory. What you explicitly ask the assistant to remember, and statements about you, your plans and your company that it derives from your conversations — each with its kind, its origin, an expiry date where one is set, and the earlier versions an entry replaced (your personal memory). The assistant also reads back in your earlier conversations where needed; nothing extra is stored for that. That way it does not start from scratch in every session. Special categories of data under Art. 9 GDPR, health information for example, go into your personal memory only when you explicitly ask for it. Under Settings, Memory you can switch off “Reference saved memories”: the assistant then uses no memories, saves no new ones and does not read back in earlier conversations.
Billing data. Subscription status, credit entries, and per request the model used, the token counts and the amount. Your payment details — card, billing address, VAT number — are collected by our payment provider in its own environment; they never reach our systems.
Log data. Request metadata such as route, status code and duration, trace identifiers and error messages. IP addresses are masked by default in our telemetry.
Support data. If you report a problem: your description, technical identifiers and — only where you tick each one separately in the form — a conversation excerpt, a screenshot or a short session recording.
Access log. When one of the provider's administrators opens the content of one of your conversations: who, when, which conversation, for what reason and with what justification.
One characteristic deserves to be named explicitly: what goes into your inputs and documents is up to you. That can include special categories of personal data within the meaning of Art. 9 GDPR — health information, for example — and equally data about third parties you write about. We do not collect such data deliberately and do not evaluate it separately.
Purposes and legal bases
Providing the service. Account and sign-in, holding conversations, working through documents, generating images, web search and the personal memory. The legal basis is performance of the contract, Art. 6(1)(b) GDPR. The memory is part of the agreed service; it is switched on by default and you can switch it off at any time under Settings, Memory. It takes in special categories of data under Art. 9 GDPR only when you explicitly ask for it. The legal basis for that is your explicit consent under Art. 9(2)(a) GDPR, which you give with that request and can withdraw at any time by deleting the entry.
Billing. Concluding the contract, processing payments, metering usage and invoicing, likewise Art. 6(1)(b) GDPR.
Protection against abuse and overload. Before each request to a model we automatically check your remaining allowance and limit the number of requests per time window. Once the allowance is used up, the request is refused until you top up or the window ends. This check decides whether the service is delivered, not anything about you as a person, and is not an automated decision within the meaning of Art. 22 GDPR. The legal basis is our legitimate interest in stable and affordable operation, Art. 6(1)(f) GDPR, together with Art. 6(1)(b) GDPR for the allowances themselves.
Operations, troubleshooting and security. Telemetry on requests and errors, capacity planning and security monitoring, Art. 6(1)(f) GDPR. We evaluate which features are used how often and how much they consume, also per user account and never on the basis of the content (Art. 6(1)(f) GDPR). To this end we store, per response, its duration, the number of model and tool calls and their outcome — never content or tool parameters.
Handling problem reports. The report itself under Art. 6(1)(b) GDPR; the optional attachments — conversation excerpt, screenshot, session recording — solely on the basis of your consent under Art. 6(1)(a) GDPR, which you give for each attachment separately and may withdraw at any time.
Operations, support and abuse investigation by administrators. The provider's administrators can see your conversations as a list — title, time, model used and size, not the content. They open the content of a conversation only with a documented reason: a support case, a concrete suspicion of abuse, a legal obligation or your consent. Every such access is logged with the person, the time, the reason and the justification; the log is part of your data export. The legal basis is Art. 6(1)(b) GDPR for support cases, otherwise Art. 6(1)(f) GDPR (legitimate interest in secure, working operation), Art. 6(1)(c) GDPR for legal obligations and Art. 6(1)(a) GDPR where you consented. We do not notify you separately of individual accesses.
Legal obligations. Retention of accounting records under Art. 6(1)(c) GDPR in conjunction with § 147 of the German Fiscal Code and § 257 of the German Commercial Code.
Without account data we cannot set up an account or perform the contract. Everything beyond that you provide voluntarily — what you write into a conversation or upload is your decision.
Conclusion and administration of contracts
Age confirmation. At your first sign-in you confirm yourself that you are at least 18 years old; we do not check identity documents. Only the result, the time and the version of the confirmation text are stored, with our identity provider ZITADEL. The application reads this information from your access token at sign-in and includes it in your data export. The legal basis is Art. 6(1)(b) GDPR, as the age limit is part of our terms of use. The information is deleted together with your account.
Declarations at purchase. For every order, that is a subscription, a change of term or a credit pack, we store what you ordered (plan, price, term), that you requested the service to begin before the withdrawal period ends and acknowledged the notice on compensation for value, the time, and the versions of these declarations, the terms and the withdrawal notice. The legal basis is Art. 6(1)(b) GDPR. We send you the contract confirmation with the terms and the withdrawal notice under Art. 6(1)(c) GDPR in conjunction with § 312f BGB. After the contract ends we keep the record until the regular limitation period expires, that is until three years after the end of the year in which your account was deleted. The legal basis for this is Art. 6(1)(f) GDPR; our legitimate interest is being able to prove your declarations in a dispute about withdrawal or compensation for value. Payment is handled by Stripe.
Cancellation and withdrawal. If you cancel or withdraw via “Cancel contracts here” or “Withdraw from contract”, we store your details (name, e-mail address, optionally a contract or customer number, the type of cancellation and, where applicable, its reason), the time of receipt, the assignment to your contract, the processing status and, where applicable, the amount of a refund. You do not need to give a reason for a withdrawal. We send the confirmation of receipt to the e-mail address given under Art. 6(1)(c) GDPR in conjunction with § 312k BGB or § 356a BGB respectively. Ending the subscription and any refund are based on Art. 6(1)(b) GDPR and are carried out via Stripe. We delete this information three years after the end of the year in which it was received; until then we keep it under Art. 6(1)(f) GDPR to be able to prove the receipt and content of your declaration in a dispute.
Processing by AI models
What you type is transmitted to an AI model and processed there — your messages, attached or uploaded documents, image descriptions, and the context the application assembles alongside them. Without that transmission there is no answer; it is the service itself. The legal basis is Art. 6(1)(b) GDPR.
We do not operate the models ourselves. We use Microsoft Azure OpenAI and Azure AI Foundry in the Sweden region. For the Anthropic (Claude) and Google (Gemini) models you can pick in the model selector, Google Cloud (Vertex AI) in an EU region is a further recipient. It also runs the model that turns text into vector representations for search: your questions, the text of uploaded documents, memory entries and passages of your conversations. Both providers process the content for us on our instructions.
Abuse monitoring at the provider. Microsoft checks inputs and AI-generated responses automatically for possible misuse; this automated check stores nothing. Where misuse is suspected, selected content may be stored for further review and viewed by authorised Microsoft staff. Microsoft’s current public documentation states no maximum retention period for this. For deployments in the European Economic Area — as ours are — this data is stored within the respective Azure geography, and the reviewers responsible are also located in the European Economic Area. The storage takes place irrespective of whether you delete the conversation with us. An exception from this storage and human review is not currently enabled for our service. The legal basis is the legitimate interest in preventing misuse, Art. 6(1)(f) GDPR. Google likewise checks for abuse with models on Vertex AI. Depending on the model and the contractual terms, suspicious inputs and responses — or, for certain models, all of them — may be stored regionally for up to 90 days and reviewed by Google staff.
No training on your content. We do not use your inputs, documents or the models’ responses to train or improve AI models. Under the contractual agreements that apply in each case, the AI providers we use do not use the inputs transmitted through our services or the responses generated to train or improve their general AI models, unless we expressly instruct or authorise such use. We give no such authorisation.
Where processing happens. We rely on models operated in data centres inside the European Union. The model selector shows for every model where it is operated. When the application picks the model for you, processing stays in the EU; the same holds for the models working in the background that nobody selects: classifying your question, searching your documents and your memory, and wording and reconciling memory entries. Where we offer models operated outside the EU, we mark them in the model selector. If you yourself choose such a model, we send your input, the necessary conversation context and any files to its operator, who processes them there — including in countries whose level of data protection does not match the EU's. The legal basis is Art. 6(1)(b) GDPR; we base the transfer on an adequacy decision of the European Commission where the provider is certified under it, and otherwise on standard contractual clauses pursuant to Art. 46(2)(c) GDPR.
What the answers are, and what they are not. Outputs are generated automatically using statistical methods. They can be wrong, incomplete or out of date, and we do not check them for accuracy. They make no decision about you and have no legal effect on you; no automated decision-making within the meaning of Art. 22 GDPR takes place.
ucando Pace: training and health data
ucando Pace is a running coach within ucando. Pace processes your training and health data only after you have explicitly consented in Pace; without consent, Pace reads and stores none of it.
What data. Activity data from files you upload (such as FIT, GPX or TCX, including exports from other services): time, duration, distance, pace, heart rate, elevation and — where the file contains it — your location track, plus the original file. Your details in the runner's passport: answers to the health check, complaints and injuries, heart-rate values, year of birth, sex, experience, personal bests, goals and races. Metrics, your level and your training plan calculated from these. Heart rate, the health check and complaints are health data within the meaning of Art. 9 GDPR.
Purpose and legal basis. Metrics such as training load, your level, your training plan and advice from the AI coach. The legal basis is your explicit consent under Art. 6(1)(a) and Art. 9(2)(a) GDPR. We record when you agreed to which version of the consent text.
AI coach. For the coach's answers we send excerpts of this data to an AI model; the providers and terms in the section “Processing by AI models” apply. We determine an activity's time zone ourselves from its starting point, without sending the location to third parties.
Withdrawal and erasure. You can withdraw your consent at any time in Pace under “Activities”, with effect for the future. We then immediately erase your activity data including the original files, your runner's passport and everything Pace calculated from them; you can export the data beforehand. If you do not sign in and upload no new data for 24 months, the consent likewise ends; we announce this 30 days in advance, with the erasure date, in the application's inbox.
Audience measurement on our websites
We count how often our public pages are viewed — the home pages, the documentation, the legal texts and the press kit —, which page or campaign visitors come from, and how often the buttons for signing up and for contacting us are clicked. That shows us which content is found and helps us improve what we offer. This measurement does not take place in the application behind the sign-in.
We use Pirsch Analytics for this (see the "Service providers" page). The measurement works without cookies and stores nothing on your device. Your browser talks only to our server; for each page view we pass on to Pirsch the address of the page, the referring page, campaign identifiers from the address, the language setting and technical details of your browser, and your IP address. From the IP address, the browser identifier, the date and a secret key, Pirsch computes a checksum to count several views on one day as one visit, derives country and city roughly from the IP address and, by its own account, does not store the IP address. We receive aggregated figures only and can neither recognise individual visitors nor link them to an account.
The legal basis is our legitimate interest in measuring and improving what we offer, Art. 6(1)(f) GDPR. You can object to the measurement at any time: on the "Cookies and local storage" page you switch it off for your browser. If your browser sends the "Global Privacy Control" or "Do Not Track" signal, we do not count your visits in the first place.
Recipients and processors
We share your data only where needed to run the service — as a rule with providers under a data processing agreement pursuant to Art. 28 GDPR; exceptions are marked in the list. Your data is not sold, and nobody passes it on for advertising.
Our providers run the data centres and the AI models, sign-in, payment processing, email delivery and the audience measurement of our websites; if you use web search or term explanations, the search provider and Wikipedia are added. Who they are — with registered office, task and place of processing — is set out on the "Service providers" page.
Beyond that, we share data where you have expressly consented, where we are legally obliged to, or where it is necessary to establish, exercise or defend legal claims.
Transfers to third countries
Our own storage is located entirely within the European Union: the application, the database, file storage, the search index and telemetry in the Azure region Sweden Central, with backups also in the EU. The AI models run at Microsoft in Sweden Central and at Google Cloud (Vertex AI) through endpoints in the EU.
Our identity provider ZITADEL is based in Switzerland; for Switzerland there is an adequacy decision of the European Commission (Art. 45 GDPR). Transfers to the United States occur in three places: Stripe also processes payment data there, web search queries may be processed there by Microsoft (Bing), and a term you have explained goes to Wikipedia (Wikimedia Foundation). Where we offer models operated outside the EU, there is the additional case that you yourself choose such a model, marked in the model selector; then your input is transmitted to the country concerned (see "Processing by AI models"). We base these transfers on an adequacy decision of the European Commission where the recipient is certified, and additionally on standard contractual clauses pursuant to Art. 46(2)(c) GDPR.
That a third country does not guarantee a level of protection comparable to the EU can be mitigated by contractual clauses but not entirely resolved — access by state authorities, in particular, remains possible.
Retention periods
Conversations and documents are kept until you delete them or your account is deleted; you can delete either yourself in the application at any time. For accounts without an active paid subscription — free accounts as well as accounts whose subscription has ended — we additionally reserve the right: if you do not sign in for twelve months, we may delete the account with all its content. We announce any such deletion at least 30 days in advance in the application's inbox and by email; until then you can export your data, and signing in or taking out a new subscription during that time prevents it.
Entries in your personal memory are kept without limit until you delete them. Only for clearly short-lived information — a trip next week, say — does the assistant set an expiry date, and it shows you that date. You can change or remove any expiry date yourself. The assistant no longer uses anything that has expired.
When you delete a memory entry, it disappears from the application at once, together with the earlier versions it replaced; otherwise a replaced version stays visible as history until you delete it. We keep deleted and expired entries marked as such for another 30 days and then delete them for good. During that time the mark stops the assistant from learning a deleted entry again from one of your conversations. When you delete a conversation, we delete what reached memory from it at the same time: the entries you asked to be remembered in it or that were learnt from it.
Telemetry log data is deleted after 30 days. The per-response details of the usage evaluation are deleted after 13 months; after that, only monthly totals per customer account remain, without details about individual users.
Attachments to problem reports — conversation excerpt, screenshot, session recording — are deleted after 30 days; the report itself is retained as a record.
Image editor jobs are deleted after 30 days.
Billing data is subject to commercial and tax retention periods. If you delete your account, those records are therefore not deleted but pseudonymised: the link to you as a person is removed, and the entries required for accounting remain and are deleted ten years after the end of the calendar year in which they arose.
With our payment provider Stripe, your customer record — name, email address, billing address and, where given, VAT identification number — and your invoices remain after your account is deleted, because we are required to retain them (Art. 17(3)(b) GDPR). We remove the link to your ucando account immediately. Ten years after the end of the calendar year of your last invoice, we delete the customer record at Stripe as well.
If you confirm in the application that you have taken note of a new version of our legal texts, we store the tenant, the document, the version, the time, your role and your user identifier as evidence (Art. 6(1)(b) and (f) GDPR). If you delete your account, this record is kept in restricted form: your identifier is replaced by a pseudonym, the entry is no longer visible in the application, and it is deleted three years after the end of the year of deletion (Art. 17(3)(e), Art. 18 GDPR).
The log of administrators' access to your content is pseudonymised when your account is deleted and erased three years after the end of the year of deletion.
Backups. Deletions take effect immediately in our live systems. Deleted data remains in our backups for up to 35 days (database) or up to 7 days (files); backups serve only to restore after an outage or error and are then overwritten automatically. If we have to restore from a backup, we delete the previously deleted data again.
Cookies and local storage
We store on your device only what is necessary to operate the application — no cookies for tracking, analytics or advertising. Our audience measurement stores nothing on your device either. A complete list of every stored item, with its purpose and retention period, is available on a separate page.
Your rights
Access and data portability (Art. 15 and Art. 20 GDPR). You do not need us for this: in the application, under Settings, Data, you will find the "Export my data" button. You receive an archive containing your profile and roles, all conversations and messages, documents including the original files and extracted text, memory, projects, image editor jobs, billing and usage, purchased entitlements, and problem reports. Payment data sits with our payment provider and can be retrieved through its customer portal; sign-in metadata sits with the identity provider. On request we will compile those for you as well.
Rectification (Art. 16 GDPR). You maintain your name and email address at our identity provider; the change is picked up at your next sign-in.
Erasure (Art. 17 GDPR). You delete individual conversations, documents and memory entries yourself. You also delete your entire account yourself: in the application under Settings, Data, using the “Delete my account” button. We carry out the deletion 14 days after your request; until then you can cancel it in the same place. That period gives you the chance to undo a misclick — the deletion itself is final and covers your conversations, documents, memory entries, projects, image-editor jobs, bug reports, your profile and your account at the identity provider. Billing records are pseudonymised rather than deleted, because we are required to retain them under commercial and tax law (Art. 17(3)(b) GDPR): your identifier is replaced by a value that cannot be reversed, the amounts remain. If you want the deletion without that period, or cannot reach your account, write to datenschutz@data-prudentia.de — then we carry it out by hand.
Restriction of processing (Art. 18 GDPR).
Withdrawal of consent (Art. 7(3) GDPR). Where we rely on your consent — for the attachments to problem reports, for special categories of data in memory and for ucando Pace — you may withdraw it at any time with effect for the future. The lawfulness of processing carried out until then is unaffected.
RIGHT TO OBJECT. You have the right to object at any time, on grounds relating to your particular situation, to processing of your personal data that we base on a legitimate interest under Art. 6(1)(f) GDPR — that covers telemetry and abuse protection. We will then stop processing that data unless we can demonstrate compelling legitimate grounds that override your interests, rights and freedoms, or the processing serves to establish, exercise or defend legal claims.
We respond to a request within one month of receipt. Where a request is extensive we may extend that period by two months, telling you within the first month and giving reasons. Handling your request is free of charge.
Right to lodge a complaint
If you consider that the processing of your personal data infringes the GDPR, you may lodge a complaint with a supervisory authority under Art. 77 GDPR, without prejudice to any other remedy.
The authority responsible for us is the State Commissioner for Data Protection and Freedom of Information of Rhineland-Palatinate (Landesbeauftragter für den Datenschutz und die Informationsfreiheit Rheinland-Pfalz). Current contact details are on their website.
You may instead approach the supervisory authority of your habitual residence or place of work — in Austria the Data Protection Authority (Datenschutzbehörde, www.dsb.gv.at).
Changes to this privacy policy
We update this privacy notice when the processing or the law changes. The version on this page applies; the date of the last change is shown above. We will inform you of material changes in an appropriate way.