Privacy policy
This policy explains what personal data the Trazvia app and the trazvia.com website process, why, on what legal basis, for how long, who it is shared with and how to exercise your rights under the General Data Protection Regulation (GDPR), Spanish Organic Law 3/2018 (LOPDGDD) and Spanish Law 34/2002 (LSSI). We do not sell your personal data.
Controller
Nesilabs Tech SL, a single-member limited company (sociedad limitada unipersonal), tax ID (NIF) B26712042, with its registered office at Majadahonda (Madrid) and registered with the Registro Mercantil de Madrid. Contact for privacy matters and for exercising your rights: hi@nesilabs.com.
We have not appointed a data protection officer for now. If we appoint one, we will publish their contact details here.
What data we process and why
- Account and sessions. You sign up with Sign in with Apple or with Google. We receive an account identifier from the provider and, if you allow it, your email (it may be an Apple relay address). We do not receive your password. We do not store your email in plain text: only a cryptographic digest (HMAC) of the verified email so that we can tell you if you already have an account; Apple relay addresses are not stored. For each session we store the platform, the device name or model, the OS and app versions and the dates of use. We use this to create and protect your account and your sessions.
- Acceptance of the legal notice and minimum age. We store the version of the notice and terms you accepted, the date and time and your statement that you are at least 14 years old. We do not ask for your date of birth or any documents.
- Location during trips. Only while a trip is active do we process your precise location (coordinates, speed, heading, accuracy and time) to detect which train you are on, estimate its position together with other riders and validate your reports. We store these points pseudonymously (linked to a random account identifier, not to your name) for 30 days at most. With no active trip, if you allow it, the app uses low-power detection (significant location changes or areas near stations) only to suggest starting a trip. With an active trip and the “Always” permission, iOS may relaunch the app in the background (significant location changes or regions) to resume tracking the trip even if you closed it; on Android, if you agree, we ask you to exclude the app from battery optimisation for the same purpose. Outside an active trip your location is not shared.
- Anonymous trajectory history. Once our system confirms the train and direction by consensus, we keep only the part of the journey made on the train, trim the sections near the origin and destination, remove user and device identifiers and merge it with other riders’ data into one consensus trace per train; information from a section is only used when there are at least 3 independent riders. From it we calculate running times and speeds per section, detect slow zones and estimate how following trains might run. You can object to this in Settings by turning off “Contribute to history”.
- Ticket. If you scan your ticket, the camera reads the code on your device and we only send its text to suggest your train. We do not store the image or the content of the ticket.
- Reports and social profile. Your train and track reports (with the point on the line where you were when reporting), free texts, confirmations and denials, display name, avatar, level, points, trips, kilometres, badges, favourite lines, reactions, “Travelling together” groups, blocks and reports about users. Others see only your display name, avatar, level and badges, never your exact position; in invisible mode, not even that.
- Device security and fraud prevention. We use Apple App Attest and DeviceCheck and Google Play Integrity to check that the app is genuine and running on a real device. We store the attestation result linked to your session, a device public key, installation and device identifiers as a cryptographic digest (HMAC) to apply the “one phone = one account” rule, your IP address only as an HMAC (never in plain text), plausibility signals from your traces and, where applicable, whether the phone is connected to a car (CarPlay, Android Auto or car audio) to rule out road trips. This prevents a single person from faking a train’s position.
- Notifications. If you enable them, we store a notification token for your device (with its platform), linked to the session, and your language. We only send you informational alerts about your own confirmed trip and your route ahead: next stop and arrival, possible delays or cancellations according to official data or the verified community, reports from other riders corroborated by several independent people, a possibly stopped train or slow section further ahead (a possible delay, never a safety alert) and works or incidents from official sources. You can turn off each type in Settings. Alerts never contain anyone’s location. The token is deleted when the session is closed or revoked.
- Bug reports and support. If you report a bug: category, description, automatic technical data (version, model, OS, language and screen, without identifiers) and an anonymised log; only if you tick the option, your approximate location (about 100 m) and a screenshot or an image from your gallery, which you cover up before sending it and which is saved as a new image without metadata. From the website, also the contact email if you provide it and an HMAC of your IP address to limit abuse. If you write to support, your message and any data you include.
Source of the data. Almost all data is provided by you or your device. Apple or Google provide us with your account identifier and, if you allow it, your email. Other users may provide data about you if they report you or if they confirm or deny your reports. Train data from Renfe and other operators is not personal data.
Automated decisions and artificial intelligence
Anti-fraud. So that no single person can fake a train’s position, each contribution goes through: (1) fixed physical plausibility rules (for example, impossible speeds, accelerations or jumps, distance from the track, consistency with the timetable and official data); (2) checks of independence between riders and of device attestation; (3) once there is enough data, our own statistical model; and (4) only in borderline cases, TypeSafe’s Jev decision model, to which we send numeric variables only (for example, account age, number of matches or share of points off the track), never your texts, your coordinates or your identifiers. The fixed rules always prevail over the models.
Possible consequences. These signals may make your contributions carry less weight or not count towards a train’s position, hold or withdraw points, hide a report or open a case for moderators. If serious plausibility failures are repeated, the system may apply a temporary 14-day quarantine during which your contributions are not used; if it happens again, it may be indefinite. You can still view the information at all times. Account suspensions for breaking the rules are decided by a person.
Your safeguards. We consider that these decisions do not produce legal effects or similarly significant effects on you (Art. 22 GDPR), because they only determine the weight of your contributions in a free, information-only service. Even so, you can ask for a person to review any automated decision, express your point of view and contest it by writing to hi@nesilabs.com or from the app. We will reply within one month.
Anthropic (Claude models). We use its API to write explanations of predictions from aggregate statistics we have already calculated and from official operator texts (for example, notices of engineering works or incidents), to answer the questions you type in “Ask Trazvia” (your question is sent together with those statistics), to summarise cases for moderators and to classify free texts (reports, complaints, display names and bug report descriptions) for moderation and spam filtering. We do not send it your location, your email or your identifiers; the texts you write may be sent for moderation. Under its commercial terms, Anthropic processes this data as a processor and does not use it to train its models. The AI never makes up train data or issues danger alerts: its texts are presented as guidance only.
“Ask Trazvia” (AI assistant)
“Ask Trazvia” is an optional assistant that answers questions about trains, stations, lines, delays, incidents and how to use the app, using data from our services. To write the answers we use the Claude models of Anthropic (Anthropic PBC, USA), which acts as a data processor.
- What is sent and to whom. We send Anthropic your question and, if you continue the same conversation, the previous ones (a few at most); the station, train or line you were viewing when you asked; and the train data our servers look up to answer you (timetables, departures and arrivals, approximate delay ranges, official incidents and notices, and grouped community figures).
- What is not sent. Not your location, your display name, your email, your account or device identifiers, or your IP address (the request comes from our servers). No data about other travellers either: the assistant cannot know where anyone is or who is on a train. Before the AI is used, an automatic filter rejects questions containing personal data (phone numbers, emails, ID documents, coordinates…); even so, please do not type personal data about yourself or others.
- It is voluntary and based on your consent. Before your first question, the app asks your permission to send it to this external AI provider. Without your permission we send it nothing, and the rest of the app works the same. You can withdraw it at any time in the app’s Settings, without affecting earlier processing (Art. 6(1)(a) GDPR). To monitor abuse, usage limits and the assistant’s security we rely on our legitimate interest (Art. 6(1)(f)).
- Retention. On our servers we keep each question in redacted form (without the personal data we detect), linked to a pseudonym of your account and with technical data (topic, sources consulted, response time, whether it was refused and your rating), for 30 days; the ongoing conversation, about 2 hours. After that only statistics without any account remain. Anthropic processes the data under its commercial terms and its processor agreement: it does not use it to train its models and keeps it for a limited period.
- Export and deletion. Your stored questions are included in “Download my data” and deleted with “Delete account”. You can also ask us to delete them sooner by writing to hi@nesilabs.com. The transfer to the USA is explained under “International transfers”.
- Answers are for information only. They are written by an AI from our data, are only a guide and may contain errors: figures are approximate (“≈”) and are not official information from Renfe or any operator. The assistant does not assess safety or warn of hazards: if you see a hazard, tell the staff or call 112. You can rate or report each answer from the conversation itself.
Legal bases
- Performance of the contract (Art. 6(1)(b) GDPR): account, sessions, tracking during the trip you start, reports, profile, points, “Travelling together”, notifications about your trip, bug reports and support, export and deletion of your account.
- Legitimate interest (Art. 6(1)(f)): fraud prevention, device attestation and the “one phone = one account” rule; service security and abuse limits using IP addresses as HMACs; building the anonymous history (you can object with “Contribute to history”); content moderation, including with the help of AI; and keeping a minimal tombstone record after a sanctioned account is deleted so that the sanction cannot be evaded. We have weighed these interests against your rights, and you can object by writing to us.
- Consent (Art. 6(1)(a)): system permissions for location, background location, notifications and the camera; the optional approximate location and screenshot in a bug report; and personalised ads (see “Advertising”). You can withdraw it at any time in your device or app settings, without affecting processing carried out before.
- Legal obligation (Art. 6(1)(c)): handling requests to exercise your rights, complying with the Digital Services Act (Regulation (EU) 2022/2065) and responding to requests from authorities.
- Minimum age: in Spain, you can consent to the processing of your data from the age of 14 (Art. 7 LOPDGDD and Art. 8 GDPR).
How long we keep data
- Account, profile, settings, trips, points, favourites, blocks, acceptances of the notice and reward requests: while you have the account. When you delete it, they are erased from the database immediately; backups are overwritten within 14 days at most.
- Pseudonymous GPS points from your trips: 30 days at most.
- Individual anonymous traces awaiting consensus: 30 days. Anonymous observations per section: 120 days. Consensus traces and aggregate statistics per section, day and hour, which no longer identify anyone: indefinitely.
- Community reports: visible until they expire; linked to your account for up to 90 days after their points are settled. They are then unlinked (without author or free text) and their votes are deleted.
- Sessions: while they are open; closed or revoked sessions, 90 more days. The notification token and the attestation result are tied to the session.
- Moderation cases and complaints: up to 12 months after closure.
- Bug reports: screenshots, 180 days; everything else, up to 24 months from the last update or until you delete the account.
- After an account is deleted: an anti-abuse tombstone record (HMAC of the Apple or Google identifier and the sanction status, with no contact details) for 90 days, or 365 if the account was sanctioned; and the now-released phone link (HMAC) for up to 12 months.
- Sign-ups per IP address (HMAC) to limit abuse: 30 days. Server technical logs, with no plain-text IP addresses or tokens: automatic rotation, 30 days at most.
Who we share data with
We do not sell or rent your personal data. Advertising is explained in the section “Advertising, in-app purchases and sponsored content”. We only share data with:
- Processors, under a contract per Art. 28 GDPR: OVH SAS (France), which hosts our servers and databases in the European Union; Anthropic (Anthropic PBC, USA) and TypeSafe (typesafe.ai), on the terms described above; Google (Firebase Cloud Messaging, Google Ireland Ltd. / Google LLC), which delivers notifications on Android (device token, alert title and text and minimal data such as the alert type or your train’s identifier); GitHub (GitHub, Inc., USA), only to manage bug reports in a private repository (description and technical data, without email or screenshot); and our email provider, for internal bug report alerts and for replying to you.
- Apple and Google, acting as independent controllers when you use their sign-in, App Attest, DeviceCheck, Play Integrity, the Apple Push Notification service (APNs, through which we send alerts directly on iOS: device token, title and text) and their stores.
- Google AdMob (Google Ireland Ltd.), as an independent controller, to serve and measure ads in the app (see the next section).
- Google Maps (Google Ireland Ltd. and Google LLC), as an independent controller: the app’s base map is displayed with the Google Maps SDK for iOS and Android. When it loads, Google receives your IP address, the map area you are viewing (position, zoom and map type), device and app identifiers and technical data, and SDK usage and diagnostic data. The dot showing your location on the map is drawn by the SDK on your device; we do not send Google your trips, your account or your precise location. More information: policies.google.com/privacy and Google Maps additional terms: maps.google.com/help/terms_maps.
- Tracks and our own map layers (railway tracks, incidents, slow sections): served by our servers; the track geometry comes from © OpenStreetMap (ODbL licence) and involves no third parties.
- Other users: they see your display name, avatar, level and badges in the on-board list and in your reports (except in invisible mode), never your exact location.
- Authorities, when required by law.
Advertising, in-app purchases and sponsored content
Trazvia is free and is funded by unobtrusive advertising: banners, native ads labelled “Ad”, a card while the train is stopped and rewarded ads that you only see if you choose to. We do not use unsolicited full-screen ads (interstitials), and advertising is never mixed with train information or notices.
- Google AdMob. serves third-party ads. Its SDK may process your IP address, the device advertising identifier (the Android advertising ID; we do not use it on iOS), technical device and app data and your interaction with ads, in order to show them, cap their frequency, measure them and prevent fraud. Google processes this as an independent controller: policies.google.com/technologies/ads.
- Consent. In the EEA, the UK and Switzerland we ask for your consent through Google’s consent platform (UMP, compliant with the IAB TCF framework) before using device identifiers or storage for advertising. You can change your choice at any time in Settings › Ad privacy. If you do not consent, you will only see non-personalised ads.
- iOS. On iPhone and iPad, ads are always non-personalised by default; we do not ask for tracking permission (App Tracking Transparency) or access the advertising identifier (IDFA).
- Location. We never share your precise location or your trips with AdMob or any advertiser or ad partner. Google may infer an approximate location (country or city) from your IP address.
- Rewarded ads. If you choose to watch one (for example, to double a trip’s points or unlock an avatar), we send Google a pseudonymous Trazvia account identifier and a single-use code so that it can confirm to us that you watched it (server-side verification). Google does not receive your name, email or location.
- Sponsored pins and results. Some businesses near stations can buy a map pin or a search result from us, always labelled “Sponsored”. We serve them from our own servers based on the map area you are viewing, without using your location or building profiles, and we only count impressions and clicks in aggregate per day, never linked to your account. We do not share any of your data with sponsors.
- “No ads” purchase. Purchases are processed by Apple’s App Store or Google Play: we do not receive your payment details, only the purchase confirmation (product and transaction identifier) to activate the benefit on your account.
- Legal bases and retention. Personalised ads and use of device identifiers or storage: your consent (Art. 6(1)(a) GDPR and Art. 22(2) of the Spanish LSSI). Non-personalised ads, aggregate sponsored counts and ad fraud prevention: legitimate interest in funding a free service (Art. 6(1)(f)). Purchases and rewards: performance of the contract (Art. 6(1)(b)). Reward requests are kept while you have the account; aggregate counts do not identify you.
International transfers
Our servers and databases are in the European Union (OVH, France). Some providers (Anthropic, GitHub, Apple, Google and possibly TypeSafe) may process data outside the European Economic Area, particularly in the United States. In those cases we rely on the EU-US Data Privacy Framework where the provider is certified (you can check this at dataprivacyframework.gov) or on the European Commission’s standard contractual clauses (Decision (EU) 2021/914), with supplementary measures such as encryption in transit and data minimisation. You can request a copy of the safeguards at hi@nesilabs.com.
Minors
Trazvia is intended for people aged 14 or over. By accepting the notice you declare that you are at least 14 years old; we do not ask for your date of birth or any documents. If you are under 14 you may not create an account or use the app. If we know, or have reasonable grounds to believe, that an account belongs to someone under 14, we will delete it. If you are a parent or guardian and believe that someone under 14 is using Trazvia, write to us at hi@nesilabs.com.
Your rights
You can exercise the following rights at any time, free of charge:
- Access and portability: in the app, Settings › Account › “Download my data” (a copy in JSON format), or by email.
- Rectification: edit your profile in the app or write to us.
- Erasure: Settings › Account › Delete account, or as explained on the Delete account page.
- Objection: turn off “Contribute to history” in Settings; for any other processing based on legitimate interest, write to us.
- Restriction of processing and human review of automated decisions (see above).
- Withdrawing your consent: in your device settings (location, notifications, camera) or in Settings › Ad privacy.
Write to hi@nesilabs.com from the email linked to your account (if you use Apple’s relay email, include your display name); we may ask you to confirm that the account is yours. We will reply within one month, which may be extended by two further months in complex cases, and we will let you know if so.
If you think we have not handled your data properly, you can lodge a complaint with the Spanish Data Protection Agency (AEPD, www.aepd.es, C/ Jorge Juan 6, 28001 Madrid) or with the supervisory authority of the EU country where you live.
Security
We encrypt communications (HTTPS), store session tokens in the device’s secure storage, rotate access tokens, keep IP addresses and device identifiers only as cryptographic digests, keep the anonymous history separate from account data, make daily backups and limit internal access to what is strictly necessary. If a security breach occurred that posed a risk to you, we would notify the AEPD and, where appropriate, you.
Changes
If we change this policy significantly, we will let you know in the app before it applies. The date of the last update is shown above.