Version 2026-10-move. Operated by SLPcentral LLC, 8735 Dunwoody Place, Ste N, Atlanta, GA 30350.
Who we are
SLPcentral is an app for speech-language pathologists, SLP assistants and SLP students. It runs on iPhone and in the browser at www.slpcentral.app. It is operated by SLPcentral LLC, a Georgia limited liability company, based in Georgia, USA ("SLPcentral," "we," "us").
This policy explains what we collect, where it is kept, who helps us run the service, how long we keep things, and what you can do about it.
The short version
Two kinds of data, kept in two places. Your account (login, profile, subscription) is kept with Supabase. Your clinical records (clients, goals, session data, notes, reports, calendar, hours) are kept in Google Cloud under a signed HIPAA Business Associate Agreement (BAA), in the United States only.
The AI never receives your client name fields. When you ask for a draft, the AI (Anthropic's Claude, under a BAA, US only) sees a random code like CLT_X92J. It can only see a name if you type one into free text, which the app warns you not to do.
Brain Dump audio is never stored. Google turns it into text and then it is discarded.
AI output is a draft. Nothing is saved until you review it and tap Save.
We don't sell your data, show ads, or use your records to train AI.
You can delete a client or your whole account from inside the app.
1. Account data (kept with Supabase)
Email address and password (the password is stored only as a secure hash by our login provider)
To create your account and let you sign in
First name
To greet you and personalize the app
Your stage (undergraduate, graduate, licensed) and onboarding progress
To show the right tools
Which version of these terms you accepted, and when
To keep a record of your agreement
How you heard about us (only if you choose to answer)
To understand how people find the app
Whether you have used the iPhone app
To decide whether to show "Get the iPhone app" on the web
Subscription status (trial, active, expired) and plan
To unlock paid features
Usage counts (for example, how many AI drafts you used today, or minutes of Brain Dump)
To apply fair-use limits. These are numbers only, never content
Grad-school application tracker entries (if you use that feature)
To show your application checklist
Your email address, if you joined our pre-launch waitlist
To tell you when the app opened
Messages you send us (in-app feedback, the website chat, the Clinical Fellowship discount form, or email to our addresses), and the reply address you give
To answer you. The CF form also asks where you graduated, where you do your fellowship and your mentor's name, so we can verify the discount
Never put client details in messages to us. Support messages and email are not kept in the protected clinical storage described below, and our email inbox is not covered by a BAA. Describe the problem without names, initials, birthdates or anything else that could identify a client.
Our login provider also keeps standard sign-in records (such as the time, IP address and device type of a sign-in) for security.
2. Clinical records (kept in Google Cloud, under a BAA)
"Clinical records" means everything in the app that is about the people you serve, plus your own training records:
Clients (each with a random code like CLT_X92J), goals and targets, session data (trial counts, cue levels, baselines), sessions, session notes (organized notes and SOAP notes you saved), progress reports, screeners and their results, and appointments on the calendar (including titles, descriptions and category names).
Your clinical and Clinical Fellowship hours, including any notes you add to them.
Client names and details, if you turn on name sync (see section 4).
Ollie's settings, note formats, memories and recent conversations (see section 6).
Where and how it is kept:
In Google Cloud, in data centers in the United States only (our Google organization is set to refuse any location outside the US).
Google has signed a Business Associate Agreement with SLPcentral, effective September 22, 2026. It covers the Google services that store and process these records.
The database has no public internet address. Only our own server program can reach it, over Google's private network.
Data is encrypted in transit (HTTPS/TLS) and encrypted at rest. Client names and details get a second, separate layer of encryption with a key that only our server program can use.
Every record belongs to your workspace. The database itself checks, on every request, that you can only see your own records.
Every time a clinical record is read, created, changed or deleted, we write an entry to an access log (who, what kind of record, when). The log never contains the record's content.
3. AI drafting and Brain Dump transcription
These features run only if you allow them. The app asks first, and you can turn them off any time under More → AI & Data Sharing. The rest of the app works either way.
Who does the work:
Transcription: Google Speech-to-Text, in our own Google Cloud project, in the United States, under our BAA with Google.
Drafting (organized notes, SOAP notes, progress reports, present levels) and the assistant Ollie (section 6): Anthropic's Claude models, through Anthropic's own service, under a Business Associate Agreement with Anthropic, with processing kept in the United States. Our requests to Anthropic go from our server program only, never straight from your device.
Brain Dump (speaking your notes after a session):
Your recording is sent to our server, split into short pieces in memory, and sent to Google Speech-to-Text.
The text comes back. The audio is then discarded: it is never saved on our servers or in our database. The copy on your phone is deleted as soon as the transcript returns, whether or not it succeeded.
We do not keep the transcript on our servers. It is used to prepare your draft.
The app does not record therapy sessions. Brain Dump is the only feature that records audio. Recording laws differ by state; make sure everyone in the room knows before you record.
Drafting:
What is sent to Anthropic: the text you dictated or typed, goal wording, trial counts and percentages, session dates and cue levels, and the client's random code.
What is never sent to AI: client names and the other client details from section 4. Our server also checks each request for anything that looks like a name or identifier and refuses it. This is a backstop, not a guarantee, so please keep names out of free-text fields.
The result is a draft. It is labeled as a draft, and nothing is saved until you review, edit and tap Save. If a draft contains a number that wasn't in your data, the app warns you.
What our AI and speech providers keep:
Neither Google nor Anthropic may use your data to train their models.
Anthropic does not keep the text we send or the drafts it returns after answering, by default. If its automated safety systems flag a request as a possible misuse of its service, Anthropic may keep that request and its response for up to 2 years to enforce its usage policy. To answer faster, Anthropic may briefly hold recent parts of a request in a short-lived cache (up to an hour), under the same BAA.
Google Speech-to-Text data logging is off (it is opt-in, and we have never turned it on).
4. Client names and details: on your device, or synced (your choice)
When you add a client, you can enter a name, birthdate, a reminder note, accommodations and plan type ("client details").
By default these start on your device. On iPhone they are kept in the iOS Keychain (encrypted by iOS). In a browser they are kept in that browser's storage for slpcentral.app on that computer. Browser storage is protected by your computer's own security, not by extra encryption from us, so use your own computer with a password and disk encryption, and don't use a shared computer.
Name sync (optional, asked once on each device). If you choose Sync, that device's client names and details are stored in our Google clinical storage (section 2), with the extra encryption layer, so they appear on all your devices. Each sync is recorded in the access log.
If you choose Keep on this device only, nothing is uploaded from that device.
You can change the choice any time under More → AI & Data Sharing. Turning sync off stops that device from sending or receiving names. It does not erase names already synced from your other devices. To remove synced names, delete the client (section 7) or your account.
Synced or not, client names and details are never sent to AI and never written to our server logs.
5. Things you type in free-text fields
Goals, notes, Brain Dumps, calendar titles and hours notes are stored in the clinical storage described above. Goal text, notes and Brain Dumps are also sent to the AI when you ask for a draft. Please don't put client names, initials or birthdates in those fields; the app warns you when it spots one. Names belong in the client's name fields.
6. Ollie, your assistant
Ollie is an optional AI assistant. It learns your documentation formats and writing style, answers your clinical questions, helps you plan sessions and draft documentation, and can prepare calendar events, hours entries, clients and goals for you to confirm.
Ollie is an AI, not a person, and says so at the start of every chat. Its answers and suggestions are for you, the clinician, to weigh. You make the clinical decisions.
Ollie runs on the same Anthropic service as drafting (section 3), under the same BAA and US-only setting. Ollie never receives your client name fields: before anything is sent, the app replaces each client's saved first and full name with their code.
Other names you type, such as a parent's, a teacher's, a nickname, a last name on its own, or a child who isn't in your caseload yet, are sent as you type them. Ollie is instructed not to repeat or save them, but please leave them out when you can.
If you ask about a client's past sessions, or when you finish a session with Ollie, Ollie may read up to your three most recent saved notes for that client (one when you finish a session). Your phone replaces your clients' names with their codes before sending them; other names in a note, like a parent's or teacher's, are sent as written. The notes are used for that reply only and aren't kept with the conversation.
Ollie's settings, formats, examples, memories and conversations are kept in the clinical storage (section 2), private to you.
Ollie may remember how you like to work (your setting, schedule, note formats and your supervisor's name for hours) and things you mention so it can check in later, like an upcoming event, without asking each time. You can see and delete these on the "What Ollie knows" screen (tap Ollie's face). It never saves moods, health details or anything about other clinicians.
Ollie saves a memory about a client only after asking you, or when you add one yourself. You can see and delete these memories on that client's profile, where you can also have Ollie forget everything about that client at once.
Ollie keeps the last 30 days of your conversations so it can pick up where you left off. Older conversations are erased automatically. You can delete a conversation yourself at any time, and conversations that mention a client are erased when that client is permanently deleted.
Ollie never saves, changes or deletes a note, logs hours, or adds or changes a client, a goal or an event on its own. It prepares a card; you confirm. Deleting a note always asks you to confirm a second time.
PDFs Ollie makes, such as resources or an hours log, are created on your phone. We don't keep a copy; they leave your phone only if you share or print them.
If you ask Ollie for a reminder, it's set as a notification on your phone only and isn't stored on our servers; a client in it appears as initials.
Ollie's weekly reminder is a fixed message scheduled on your phone and never contains client information.
When Ollie finishes a reply or a note while you're elsewhere in the app, your phone may show a short notice ("Ollie replied." or "Your note is ready to review."). It never names a client or quotes a note, and it appears only if you allowed notifications.
7. How long we keep things
Clinical records
Until you delete them. We never delete records because an account is inactive.
A client you delete
Moves to Recently Deleted for 30 days, where you can restore it. After 30 days it is permanently erased, with everything attached to it (goals, data, notes, reports, screener results, synced name and details). You can also choose Delete now to erase it right away.
A note you delete
Notes you delete go to Recently Deleted for 30 days. You can restore them there; after 30 days they're erased for good.
Your account, when you delete it
Your clinical records, synced client names and assistant data are erased immediately. Your login, profile, subscription record and usage counts are erased at the same time.
Backups of the clinical database
Google keeps automatic backups for about 30 days (plus 7 days of point-in-time recovery). Deleted records can remain in these backups until they age out, then they are gone. We only use backups to recover from an outage or error.
Access log (who opened which record, and when; no content)
At least 6 years, as health-privacy recordkeeping. It is kept even after the records it refers to are deleted, and identifies you only by a random account ID.
Google's own security logs of admin and data access (no content)
About 30 days today.
Text sent to Anthropic (drafting and Ollie)
Not kept after the answer by default, apart from a short-lived cache of up to an hour; flagged requests up to 2 years (section 3).
Ollie conversations
The last 30 days; older ones are erased automatically. You can delete a conversation yourself at any time, and conversations that mention a client are erased when that client is permanently deleted (section 6).
Support messages and email to us
Deleted within 24 months. They are not deleted automatically with your account, because they are not linked to it; ask us and we will delete them sooner.
Account-data backups
Our account database's current plan keeps no automatic backups. Any manual backup we make is kept encrypted on our own computer, never in cloud storage, and deleted within 30 days.
Records from before the move
During the move to Google, a copy of clinical records remains in our previous database, and in an encrypted offline backup, as a safety net. We delete both within 30 days after the Google copy is verified.
Billing records
Kept by Apple, RevenueCat and (for web purchases) Stripe as long as they require for tax and accounting.
8. Who helps us run SLPcentral
We use a small number of service providers. Each gets only what it needs to do its job. The full list, with what each receives and where, is in our Subprocessors list (More → Subprocessors). In short:
Google Cloud: clinical records and Brain Dump transcription (under a BAA).
Anthropic: AI drafting and the assistant Ollie (under a BAA, US only). Receives text with client codes, never client name fields.
Supabase: logins and account data.
Apple and RevenueCat: subscriptions and purchases. RevenueCat receives your random account ID, not your name or email.
Resend: sends account emails (confirm your email, reset your password, welcome).
Vercel: hosts the website. It also counts page visits on our public marketing pages only, without cookies; pages you use after signing in are not counted.
Sentry: reports crashes in the browser version only, with your random account ID. It is set up not to record your screen, what you click or type, or console messages.
Expo: delivers app updates to your iPhone.
Cloudflare: runs our domain name and forwards email sent to our slpcentral.app addresses (such as privacy@slpcentral.app and support@slpcentral.app) to our inbox.
Google (Gmail): the inbox where that email arrives. It is an ordinary email account, not covered by a BAA.
Telegram: sends the owner a notice that a new message arrived (which form, whether a reply address was left). It never receives the message itself.
Only Google (for storage and transcription) and Anthropic (for AI drafting) have signed BAAs with us, and only they receive clinical records. The others are set up not to receive clinical records. That is why you must never put client details in support messages or email.
We will update the Subprocessors list and tell you in the app at least 30 days before a new provider starts receiving clinical records, except when we must change a provider urgently to keep the service secure or available; then we tell you as soon as we can.
We do not sell or rent personal information, share it for advertising, or use it to train AI models.
9. Who at SLPcentral can see your data
SLPcentral is run by one person. The owner has administrator access to the systems. The owner opens clinical records only when you ask for help in writing, to investigate a security incident, or to restore service after an outage, and only what is needed; each time is logged. Clinical storage access is also recorded in Google's audit logs. We do not read your clinical records to improve the product; product testing uses made-up data.
We may disclose information if the law requires it (for example, a valid court order). If that ever involves your records, we will try to tell you before we respond, where the law allows and time permits.
10. Your responsibilities as a clinician
You decide what goes into SLPcentral, about whom, and why. The people you serve are often children, and their information is usually protected by HIPAA, FERPA, state law, your employer's policies or your university's rules. You are responsible for following them, including getting any consent you need before recording or using AI, and for keeping required records in your official system (for example, your district's IEP system or your practice's records system).
Business Associate Agreements. If you are a HIPAA covered entity (for example, a private practice that bills insurance electronically), HIPAA requires a Business Associate Agreement between us before you store patient information in SLPcentral. We offer one. Email privacy@slpcentral.app with your practice name and we will send it for signature. Do not include patient information in that email. If you work in a school, ask your district whether it needs its own agreement with us. If you are a student clinician, your university clinic or externship site may own your clients' records; follow its rules.
Student information. Many of the people our users serve are K-12 students. We use their information only to provide SLPcentral to the clinician who entered it. We never sell it, use it for targeted advertising, or build profiles of students. We delete it when the clinician asks, or on a verified request from the clinician's school or district, after telling the clinician.
11. Children
SLPcentral is for adults (18+). We do not knowingly collect information from children. Information about child clients is entered by their clinician, not by the child, and is handled as described above. SLPcentral is not designed for children and must not be used by them.
12. Your choices and rights
You can, at any time:
See and correct your records in the app.
Download or share what the app already exports: print or share a client's session data and reports, copy your notes, and export your hours. A one-tap export of everything isn't available yet. If you need a copy of something the app can't export, email privacy@slpcentral.app and we will help.
Delete a client (Recently Deleted, then permanently after 30 days, or Delete now) or your whole account (Profile → Delete account). Deleting your account does not cancel an Apple subscription; cancel it first in iPhone Settings → Subscriptions.
Turn off AI drafting and transcription, or name sync, under More → AI & Data Sharing.
Ask what we hold about you, or ask us to delete support messages.
These rights apply to everyone, wherever you live. We will not treat you differently for using them.
If you are the parent or guardian of a client, please contact your child's clinician or their school or practice. They control the records. We will help them respond.
Do Not Track. We don't track you across other websites, so there is nothing for a browser "Do Not Track" signal to turn off. We respond to it the same way either way.
13. Face ID and Touch ID
If you use Face ID or Touch ID to unlock the app, the check happens entirely inside iOS. The app only learns "success" or "failure." We never receive or store face or fingerprint data.
14. Security
In addition to what is described in section 2: the app locks after 2 minutes of inactivity; sign-in tokens are kept in the iPhone's secure storage; our Google organization blocks public database addresses, public storage and downloadable service keys; each part of our server has only the permissions it needs; the key our server uses to reach Anthropic is kept in Google's secret storage, never in the app; and every administrator account we use has two-step verification.
No system is perfectly secure. If a breach affects your information, we will notify you as required by law.
15. Changes to this policy
When we make a material change, we will update the version date and tell you in the app before it takes effect. Where the law or the change requires it (for example, sending a new kind of data to a new place), we will ask for your permission again.
16. Contact
Privacy questions and requests: privacy@slpcentral.app. Help with the app: support@slpcentral.app. Please never include client details in email. Mail: SLPcentral LLC, 8735 Dunwoody Place, Ste N, Atlanta, GA 30350.
This policy is written in plain language and in good faith. It is not legal advice.