Privacy
Nichecast stores the minimum it needs to run your channels:
- Account — your email address, used for one-time sign-in codes and for the email telling you a video has finished or could not be finished. That finish email is switched off under Account & connections → Notifications; the sign-in code cannot be switched off, because it is how you sign in. Payment receipts come from PayPal, on PayPal's own account — Nichecast sends none. No passwords are stored.
- Channels & videos — your channel briefs, generated scripts, narration audio, and rendered videos, kept so you can inspect and retry jobs.
- YouTube connection — an encrypted refresh token, obtained through Google sign-in when you connect a channel, that lets Nichecast upload to your channel. It is used only for uploads and is deleted when you disconnect the channel.
- Billing — subscriptions and credit-pack purchases are processed by PayPal; Nichecast never sees your card or bank details.
- Product usage measurements — which steps of the product are used (a page view, an idea-tool run, a sign-up, a channel, a video, a checkout), so we can see where people get stuck. This is cookieless and does not follow you between visits; the section below says exactly what it stores.
- Support chat — if you send a message in the chat, the conversation, who it belongs to and the IP address it came from; the support chat section says what, for how long and who receives it.
We don't sell your data or use it to train models.
Deleting your account
You can delete your account yourself, from the app: open Account & connections under the studio title, then Delete account. You'll be asked to type a word to confirm, because it is immediate and cannot be undone.
It removes the account and everything we hold for it — the channels you described, every video job and its rendered files, the stored access to your connected YouTube channel (which we also withdraw at Google), your billing records and credit history, your support-chat transcripts, and the single row that counts return visits. Videos already uploaded to your YouTube channel are not touched: they are on your channel and are yours to keep or delete there.
If you have a live subscription, cancel it first — deletion is refused while one is running, whether you do it yourself or ask us to do it for you. Deleting removes our record of the subscription but does not stop PayPal, so it would keep charging for an account that no longer exists.
If you can't sign in, email support@tiraisoft.com and we'll do it for you.
How we measure product usage
This describes what the code actually stores, rather than a general statement of intent.
- No cookies, and no identifier that survives your visit. Each visit gets a random id held in your browser's
sessionStorage. It is deleted when you close the tab, so two visits cannot be linked to each other or to you. - Measurement never stores your IP address. To count distinct people per day we hash your address together with that date and a secret unique to Nichecast, keep the first 8 bytes, and discard the address. Because the date is inside the hash it cannot be matched to another day, and because the secret is unique to this product it cannot be matched to another Tiraisoft product.
- Only named steps are recorded. An event has to be on a fixed list we declare in the code. We do not record the briefs or topics you type, and page addresses are stored without their query string, except a short campaign label such as
?src=tiktok-archivewhen a link we published carries one — a label we chose, counted per day, never tied to a person. - Referrers are reduced to the site name. If you arrive from a search engine we store the host (for example
www.google.com) and never the search terms. - Device and system are kept as broad buckets, never as the browser's own string. We record the kind of device (phone, tablet, desktop, or an automated crawler) and the operating system with its major version only —
iOS 17, neveriOS 17.4.1. The browser'sUser-Agentstring is read once to work that out and then discarded, because the full string is detailed enough to single somebody out. The network a visit arrives from is read the same way, only to recognise the automated link scanners that hosting and email-security companies run, and then discarded: what is kept is the crawler verdict, never the network. - Country and region come from Cloudflare, not from an address we keep. Cloudflare works out the country at its edge, so this measurement gets the answer without ever seeing or storing the address that produced it.
- These are counted separately and never combined. We keep a daily total per device kind, and a separate daily total per country, and never a row that joins them — so a total can never be narrowed down to "the one person on a tablet in Iceland". It also means we cannot ask which system a particular country's visitors run, which is the trade we chose.
- We measure how long a page was visible. The clock stops whenever you switch away from the tab, so a page left open in the background is not counted as time spent reading.
- It stays in our own database. These measurements go to no analytics company. The individual rows are deleted after 30 days; what remains is a set of daily totals — per step, per device kind, per system, per country, per region, and per page — none of which can be traced back to a visit. The one exception is the account row described next.
- If you are signed in, we count return visits using your account — and nothing else. We keep one row per account holding the first and the last day that account was active. No pages, no events, no times: that row cannot say what you did, only that you were here on those two days. It exists so we can tell repeat use from one-off curiosity, and it uses the account id we already hold rather than a cookie or a device fingerprint. Signed-out visits carry no such id at all and are counted only as an anonymous total. This is the one measurement kept for longer than 30 days, because a return after five weeks would otherwise read as a new person; it is deleted when your account is deleted.
Free tool
The free idea tool is different: to enforce its daily limit it keeps the IP address of the connection, as the raw address, with a count of how many times it was used that day. That record expires on its own after about a day (26 hours). Before a request is answered, the tool also runs a browser check to keep automated abuse out, and the IP address is sent with it to Cloudflare, which runs the check.
Support chat
Everything above describes measurement. The support chat is different, because it exists to keep a conversation, and none of the measurement's properties apply to it. It is offered on every page, signed in or not. If you send a message in it, we store:
- the transcript — every message you send and every reply;
- who it belongs to — your account id if you are signed in, otherwise a random visitor id that the chat creates when you send your first message and keeps in your browser's
localStorage, so the conversation can be reopened on a later visit; - the IP address of the connection the conversation started from, stored with the conversation and used to limit abuse of the chat.
Each conversation is classified automatically (for example as a question, a bug report or a feature request). A short summary of a bug report or a feature request is emailed to our own feedback inbox so the team sees it.
Chat conversations are kept until they are deleted; they have no automatic expiry. A signed-in account's conversations are deleted when the account is deleted. A conversation held under a visitor id is not tied to an account, so to have one removed, email support@tiraisoft.com.
Who else touches your data
These are categories of recipients, not a list of names. Article 13(1)(e) GDPR asks for the recipients or categories of recipients, and the category is the disclosure: a name here would go stale the day a provider changed, and this page would then be confidently wrong about where your text goes.
- A third-party large-language-model provider — the text you submit, meaning your channel brief and the scripts and image prompts derived from it, is sent to it so scripts, narration text and thumbnails can be produced. Your support chat messages are sent to it too, so it can write the reply and classify the conversation.
- A third-party text-to-speech provider — the narration script is sent to it to be spoken.
- A stock-media library — search terms taken from your script are sent to it to find footage.
- A cloud compute provider — rendering runs on containers we rent; the audio, images and footage for that one job are processed there while it runs.
- Cloudflare — hosting and the database, the free tool's browser check (Turnstile), which receives the IP address of the connection, and delivery of the support chat's bug-report and feature-request summaries to our feedback inbox.
- A third-party payment processor — subscriptions and credit packs are paid for on its own site, so the card or account details you enter go to it and never reach us. It receives the plan or pack you picked and an internal id that links the payment back to your account, and it sends back the subscription's status or the payment's outcome. As with every entry here, the category is the disclosure: Article 13(1)(e) GDPR asks for the recipients or categories of recipients.
- Google / YouTube — uploads to the channel you connected, as set out below.
Google user data
Nichecast uses YouTube API Services. Nichecast's use and transfer of information received from Google APIs adheres to the Google API Services User Data Policy, including the Limited Use requirements. The Google Privacy Policy also applies to data Google processes. You can revoke Nichecast's access to your YouTube account at any time by disconnecting the channel in the app or via your Google account security settings.
Contact
Nichecast is a Tiraisoft product. Questions about this notice or your data: support@tiraisoft.com.