Interact
'Where do my participants' answers go?' A DPO asks that question professionally, a consultant under NDA asks it out of caution, and many trainers never ask it — until the day an institution asks them. With most online tools, the honest answer is: on the vendor's servers. With SnapInteract, the answer fits in one sentence: your data stays on your Mac. This article explains how that is technically possible, and what it changes in practice.
The Mac itself hosts the participant page
This is the heart of the architecture, and it is counter-intuitive if you come from online tools: when you start a session, your Mac becomes the server. It hosts the web page your participants open. The projected QR code — or the short six-character code — points to your machine, on the room’s local network. Any browser will do: no app to install, no account to create.
When a participant votes, answers a quiz or posts an idea, their answer travels from their phone to your Mac over the room’s network. It does not need to cross the internet, and it does not land in a database on the other side of the world. We describe the participant side in joining a session in 3 seconds.
The question is not 'can this vendor be trusted', but 'why should a vendor be needed for a room to vote'.
A local file, automatic saving, a backup copy
Once the session ends, where does the data go? Nowhere: it is already there. Your courses, sessions and participant answers live in a file on your Mac, with automatic saving and a backup copy handled by the application. We keep nothing on our servers — not as a contractual promise, but by construction: there is no SnapInteract database where your content would be filed.
That local file hands you the whole data life cycle. Want to keep a year’s history? It is on your disk, backed up with the rest of your machine. Need to purge a group’s answers at their request? You delete on your side, and it is done — there is no copy elsewhere to have erased. Want to pass on a debrief? The PDF export of the statistics and the zip of visuals are generated locally, when you decide.
The optional internet relay: carrying the session, not keeping it
One real-world case remains: participants’ phones are not always on the same network as your Mac — a walled-off guest network, participants on mobile data. For that case, an optional internet relay exists. Its role is deliberately minimal: it carries the session between the phones and your Mac, like a pipe, without storing the data at rest. Answers pass through, arrive on your Mac, and that is where they live.
The distinction between carrying and keeping is exactly the one a DPO cares about. A service that keeps named answers becomes a processor under GDPR, with a register, retention periods and access rights to organise. A relay that carries without keeping drastically shrinks that surface: durable storage stays with you, on your machine, under your direct responsibility.
No participant account: data you never collect needs no protecting
The best personal-data processing is the one that never happens. Your participants never create an account: no email, no password, no profile growing session after session at a third party. They scan, they take part, full stop. What you choose to collect — a first name for tracking, anonymous answers for a sensitive vote — remains a pedagogical decision that belongs to you, not a platform requirement.
It should be said plainly: a tool that requires an account per student creates, by construction, a register of people at a third party, with everything that implies about consent, informing families and retention periods. A tool with no participant account creates nothing of the sort. That is a difference in kind, not in degree.
- No participant account: no email, no password, no third-party profile
- Answers travel from the phone to the Mac, over the room's network
- Courses, sessions and answers in a local file, with a backup copy
- Optional internet relay: it carries the session without storing data at rest
- Deletion under your control: erasing on your side is enough, there is no copy elsewhere
AI and analytics: the two awkward questions, answered frankly
Two topics deserve a straight answer. First, artificial intelligence: SnapInteract offers a synthesis of participant feedback, and it runs locally, via Apple Intelligence, on Apple Silicon Macs with a recent macOS. Nothing is sent to an external service for analysis; and even without AI, a simple summary always exists. How it works is detailed in local AI feedback synthesis.
Second, the application’s own audience measurement: yes, it exists — we need to know which features are used in order to improve the product. It is aggregated, hosted in the European Union, contains no personal data, and can be turned off in the settings. Your teaching content and your participants’ answers are never part of it.
What it changes for a consultant under NDA, an institution, a DPO
For the consultant under a non-disclosure agreement — facilitating at a client’s, with real cases and sensitive figures in the deck — the equation becomes simple: decks and answers stay on their machine, and there is no extra cloud service to get approved by the client. For an institution, the impact assessment shortens accordingly: no student account created at a third party, no outsourced answer database, an audience measurement that is aggregated, EU-hosted and switchable off.
For the DPO, finally, the architecture pre-answers the processing register’s questions: where is the data (on the facilitator’s Mac), who accesses it (the facilitator), for how long (as long as they keep their file), what does the relay do (it carries without keeping). No tool replaces rigour — informing participants and framing what you collect remain good practice — but starting from a data-minimising architecture changes the conversation. To see the whole product built on that principle, the product page gives the full tour.
Frequently asked questions
Do participant answers go through your servers?
Not by default: the Mac itself hosts the participant page and answers travel over the room's local network. If the phones are not on the same network, an optional internet relay carries the session without storing the data at rest.
Do my students need to create an account to take part?
Never. They scan a QR code or type a short six-character code and land on the activity page in their browser. No email, no password, no profile created at a third party.
Does the AI feedback synthesis send data to an external service?
No. It runs locally via Apple Intelligence, on Apple Silicon Macs with a recent macOS. Nothing is sent to an external service, and a simple summary always exists even without AI.
What should I tell my DPO or my institution?
The essentials fit in four points: data lives in a local file on the facilitator's Mac; there is no participant account; the optional relay carries without keeping; audience measurement is aggregated, EU-hosted, free of personal data and can be turned off in the settings.
