Work/Life 2.0

Privacy

What happens to your data

Who receives your data, why we use it, how long we keep it, and how to contact us. Swiss data protection law applies.

Who is responsible

Olaf Kupschina, in Switzerland, who runs Work/Life 2.0. There is no company behind the lab, so he is the one responsible. Write to olaf@paradaice.info about anything on this page.

This website

The site counts visits with Cloudflare Web Analytics (Cloudflare, Inc., USA). It sets no cookies and does not identify you; it records the page, the page you came from, your country, your browser and device type, and how fast the page loaded. Apart from that one script, the site's security policy blocks scripts from anywhere else; only the showroom pages run the site's own two small scripts. Fonts and images are served from this site. Cloudflare also hosts the website and handles your IP address to deliver pages and protect the service. The "Watch it live" link on the front page and on the explained page opens a dashboard hosted by Grafana Labs (USA); the site itself makes no requests to it.

When you ask for a seat

With the form on this site, the lab receives your name, your email address and what you wrote about your idea. If you email join@ instead, it receives your email address, sender name, subject and message. We use these to review your request, respond and, if approved, set up your seat. We do not sell application data or add you to a mailing list.

Your application is handled in these places:

The intake parser and acknowledgement use fixed rules. A person reads every request and decides. Information read by Claude agents is passed to Anthropic (USA) for processing; the acknowledgement alone does not approve a seat.

If you get a seat

Your lab address forwards to your own inbox, so that inbox's address is kept at ForwardMX and in the lab's database. Your name, lab address, GitHub username and project name go into the lab's member list, which other members can read. GitHub (USA) receives your GitHub username for membership and project access. So that you can sign in to Claude with GitHub, your name, lab address and GitHub username are also kept on the lab's own sign-in server, rented from Hetzner in Germany. Anthropic (USA, the makers of Claude) receives your name and lab address when you first sign in to Claude, which creates your account there. These services also process the content you submit when you use them, under their own terms and privacy notices.

Usage reporting

Claude Code on your seat reports usage figures to the lab's own monitoring. The monitoring runs at Grafana Cloud (Grafana Labs, USA). The data is stored in Grafana's Switzerland region.

Only counts and amounts are sent: the number of sessions, tokens used, cost, lines of code changed, commits and pull requests created, active time, which model was used, and the account it came from. That account is the email address and account id Claude knows you by.

These are not sent: what you type, what Claude answers, the contents of your files, the commands that are run, and tool output. Claude Code leaves these out by default, and the lab does not switch them on.

The lab's AI gateway, the service tiny-claw workers use, also records tokens and calls per seat. The lab reads per-seat spend from OpenRouter (USA) and Anthropic (USA) for the same purpose. OpenRouter receives the AI requests the lab's own workers send.

Only the lab's administrators see figures for one seat. Public dashboards show totals for the whole lab, never anything about one member.

The lab does this to keep seats free by knowing what AI costs, to see whether seats are used, and to notice problems early.

Grafana keeps these figures for 14 days. Longer-term numbers are kept only as lab-wide totals.

If you object, write to olaf@paradaice.info. Your seat's figures are then left out of the dashboards.

The lab chat

The lab operates its own Zulip chat on a server rented from Hetzner in Germany. Access is restricted to members and their authorised agents.

The server stores your account name, an email address for you, and your messages. People and agents with access to a channel can read its messages; direct messages are available to their participants. The lab administrator also has technical access to data on the server.

If someone reports a problem, the lab may read the messages involved, including private ones, to find out what happened. Members cannot delete messages they have sent (only the chat's administrators can), and an edited message keeps its earlier version, so the record stays complete.

Messages read by Claude agents are passed to Anthropic (USA) to generate responses. Storing chat history on the lab's server does not keep this AI processing on that server.

You sign in with GitHub, so the chat learns your GitHub username and the email addresses GitHub confirms for you. It never sees your GitHub password.

Your chat account is created at the address you wrote to us from, because that is the address GitHub already confirms for you. That is how signing in with GitHub recognises you without you setting anything up. So your own inbox address is also stored on the chat server. Members who joined before 28 September 2026 have an account at their lab address instead, and keep it.

If you leave, your account is switched off and your messages stay where they are, so the conversations still make sense to the people who were in them. You can request deletion using the contact below.

Members under 18

The lab admits people from 13. Everyone under 18 needs a parent or guardian to say yes before a seat is made. The form asks your age group. If you are 13 to 17, it also asks for a parent's or guardian's name and email. We email them one link to confirm. We cannot check an ID, so we rely on what you tell us, on that confirmation, and on a person reading every request.

For a member under 18 we also keep: their age group (not their date of birth), their parent's or guardian's name and email address, and the date they confirmed. We use these to ask for consent, to show that we have it, and to be able to reach a parent if something goes wrong. We keep them while the seat exists. When the member leaves, or when a parent asks, we delete them. A parent or guardian can see, correct or delete their child's data and can withdraw consent at any time. Withdrawing consent ends the seat. Write to safety@paradaice.info.

The agents young members work with are AI (Claude, made by Anthropic), not people, and they follow extra safety rules for members under 18. How the lab keeps young members safe, and how to report a problem: Young members.

Which laws apply

The lab is run from Switzerland, so the Swiss Federal Act on Data Protection (FADP) applies. Where a member lives in the EU or EEA, the GDPR applies as well. Article 8 of the GDPR lets a child consent alone from an age between 13 and 16, depending on the country. We ask for a parent's or guardian's consent for everyone under 18, which is stricter than that limit in every member state.

We do not admit children under 13 and do not knowingly collect their personal data. The form stores nothing when someone says they are 12 or younger. If we learn that a member or applicant is under 13, we delete their data. The US Children's Online Privacy Protection Act (COPPA) covers children under 13, and we comply with it by not admitting them.

How long, and your rights

There is currently no automatic deletion schedule for application emails, account records or chat history. This includes declined applications and records of former members.

You can ask to see the data we hold about you, have it corrected or request deletion. We will explain what can be removed, any limits that apply, and whether a provider also needs to act. Write to olaf@paradaice.info.

Back to the lab's page

Last changed 28 October 2026.