Privacy Policy

Contents.

1. Who we are and who this policy covers
2. What we collect
3. The Prepara Course Workspace
4. Google API Services and other connected file sources
5. AI features
6. How we use information
7. Learning analytics, research and our own models
8. Cookies, browser storage and website traffic
9. Who can see your information
10. Service providers we use
11. Student records, FERPA and state student privacy laws
12. Learning-management systems (LTI)
13. Security
14. How long we keep information
15. Your privacy rights and choices
16. Where your information is processed
17. Younger users
18. Changes to this policy
19. How to contact us

1. Who we are and who this policy covers

This Privacy Policy explains how Prepara LLC (“Prepara,” “we,” “us”), an Oklahoma company, collects, uses, shares and protects personal information. It covers the Prepara courseware platform and its books at app.preparacourse.com, our public website at preparacourse.com, and the tools that run with them: the AI tutor and study tools, the in-class response activities, Peer Evaluations, the Evaluation Studio, the Selling Simulation, the Sales Management Simulation, and the Prepara Course Workspace. In this policy we call all of these together the “Service.” Payments are processed by Stripe on behalf of PhD Analytics LLC, so a card charge appears as “PhD Analytics.”

This policy applies to everyone who uses the Service:

• Students, whether you reach Prepara through your school, through your learning-management system (LMS) such as Canvas, or by buying access yourself.
• Instructors and teaching staff, including teaching assistants and co-instructors, and instructors who use the Course Workspace.
• School staff, such as administrators and IT staff who set up Prepara for a school.
• Visitors to our public website who do not have an account.

When a school uses Prepara under a signed agreement with us that covers student data, that agreement controls if it differs from this policy. This policy does not cover other companies' websites or services, including your school's LMS, even when you reach them through a link in Prepara. Their own privacy policies apply there.

2. What we collect

We collect only what we need to run the Service and to improve it. Some of it you give us, some is created as you use the Service, and some comes from your school or its LMS.

Information you give us

• Account details: your name, your email address (a school-issued academic email is required to create an account), and your password, which we store only as a secure hash. You may turn on two-factor authentication or add a passkey; for a passkey we store only its public key, which cannot be used to sign in as you. If you use “Sign in with Google,” Google sends us your name, email address and profile picture.
• Instructor details: if you ask for instructor access, your school, the courses and sections you teach, and your role.
• Content you create: your notes, highlights, voice notes and their transcripts, canvas boards, Checkpoint Quiz answers, assignment submissions, answers in in-class activities, ratings and comments in peer evaluations, and recordings you make in the Evaluation Studio or the simulations.
• Survey answers: your answers to the optional confidence and AI-literacy questions.
• Your research choice: whether you have turned off the inclusion of your de-identified learning data in published education research. This is a single on-or-off setting in Settings, and it is on unless you turn it off.
• Your agreement to our terms: the date you agreed to the Terms of Service and this Privacy Policy, which version you agreed to, and whether you were signing in on our site or from your school's LMS.
• Payment records: an order record with the amount, the date and the status. Stripe handles the card itself, and we never receive your card number.
• Messages to us: what you write to our support team, to the AI support assistant, or through a contact form. A contact form also records the network (IP) address and browser details it arrives with, so we can answer and prevent abuse.

Information created as you use the Service

• Learning activity: the chapters, lessons and videos you open, how far you scroll or watch, how much active time you spend, which study tools and learning widgets you use, your answers and results on checkpoints, quizzes and assignments, the questions you miss and later recover, and your progress.
• AI interactions: the text or audio you send to AI features and their replies, including your AI tutor conversations.
• Device and connection details: the kind of device you use (phone, tablet or computer), its operating-system family, screen size, whether you use Prepara in a browser tab or as an installed app, and a random install identifier kept in your browser (explained in the section on cookies and browser storage). When you sign in, your session record keeps the network (IP) address and browser details it started from, and our security logs keep the same details for sign-ins and administrative actions.
• Coarse location: a county-level location recorded with your learning activity, described in section 7.
• How you arrived: whether a session started inside your LMS or directly on our site, so we can tell how Prepara is used inside a school's LMS.
• Notifications: if you turn on browser notifications, the delivery address your browser's push service gives us for your device.

Information from your school and its LMS

• When you launch Prepara from your LMS: your LMS sends us your name, email address, your role in the course, the course and section, and the LMS's own identifiers for you and for the course, so we can sign you in to the right section.
• Rosters and groups: for a section an instructor has connected, the list of people enrolled and their roles, and, if the instructor allows it, the course's groups and their members.
• Enrollment details from your school: course numbers, section numbers, term dates and similar details your school or instructor uses to set up the class.

Voice, video and recordings

• Voice notes: when you record a voice note, we turn it into text for your own use.
• Evaluation Studio and simulations: some courses use the Evaluation Studio or a simulation for practice, such as a recorded sales conversation or a live discussion room. When you take part, we record the video and audio of that session, turn the speech into a transcript, and may produce AI feedback on it. The transcript may mark who said what within the recording.
• What we do not do with recordings: we do not use your voice or your face to identify who you are, and we do not build a voiceprint or a faceprint from any recording.

What we do not collect

• We do not ask for or store demographic information about you, such as race, gender, age group or income. A school that wants to study outcomes by group does that in its own systems by matching our reports to its own records; we never hold that data.
• Our learning-activity records never contain the text of your work. They store counts, timings and short labels, never your notes, your answers, your tutor messages, your voice-note transcripts or your board drawings. Your work itself is stored with your account, separately.
• We never see or store your card number.
• We do not track you across other websites, and we never sell your personal information.

3. The Prepara Course Workspace

The Prepara Course Workspace is a tool for instructors. An instructor connects their Canvas course and, if they choose, the place where they keep their course files, and then asks the Workspace to find problems, answer questions about the course, and prepare changes. This section explains what the Workspace can reach, what it keeps, and how an instructor disconnects it. Students do not use the Workspace directly, but it can read information about students in the instructor's own Canvas course, so this section matters to students too.

The Canvas connection

• How it connects: the instructor creates a personal access token in their own Canvas account and gives it to the Workspace. The token lets the Workspace act in Canvas with that instructor's permissions, and no more.
• How the token is protected: we encrypt the token before storing it, using AES-256-GCM with a separate key for each connection. That key is itself encrypted by a master key that exists only as a secret inside our server environment, never in our code or our database. The token is decrypted only in memory, at the moment a Canvas request is made, and it is never written to a log, an error message or a report.
• What it is used for: the Workspace uses the token to read the instructor's course and to make only the changes the instructor approves. Nothing is written to Canvas until the instructor reviews a preview of the change and presses Apply. There are no background changes. This includes grades: the Workspace can post scores or feedback to a Canvas assignment only as a prepared change the instructor reviews and applies, and it never changes grades that Prepara's own books send through the LMS connection.
• What it reads and keeps: the course structure (modules, pages, assignments, quizzes, discussions, files and their dates), a copy of the course files the instructor works with, and the course's people list: each person's name, role, section, enrollment status, Canvas login, when they were last active in the course, and how much time they have spent in it. Student submissions and grades are read when the instructor asks about them. Grades are read live from Canvas and are not stored as a separate record. If an instructor records an accommodation such as extra days or extra time, the Workspace keeps only the student's Canvas identifier and the numbers, never a description of the reason.

Connected file sources

An instructor may also connect a place where they keep course files. Each connection is the instructor's own account, which they approve on the provider's own screen and can remove at any time.

• Google Drive: the Workspace asks Google only for the drive.file permission (https://www.googleapis.com/auth/drive.file). With it, the Workspace can see only the files and folders the instructor picks in Google's own file picker. It cannot see anything else in the instructor's Google Drive, and it does not change the instructor's Google Drive files.
• Choosing course folders: after an instructor connects Dropbox, OneDrive or Box, the Workspace lists the names of their folders, not what is inside them, so it can suggest which folders go with which course. It then reads, and for Dropbox writes, only the folders the instructor confirms for each course. A course can have several folders.
• Dropbox: the instructor confirms the folders that hold each course. The Workspace reads the files in those folders. It creates, changes, renames, moves or deletes files there only when the instructor has reviewed and applied that change, and it checks the result afterward.
• OneDrive and Box: Microsoft and Box do not offer a permission limited to one folder, so the permission the instructor grants lets the Workspace read their files. The Workspace reads and keeps only the folders the instructor confirms for each course, and it cannot write to OneDrive or Box at all.

What we store from these sources: the names of the instructor's folders used to suggest course folders, and the names, folder paths, sizes, dates and content fingerprints of the files in the confirmed folders or picked items, a copy of each file's contents, and the text we extract from them, so the Workspace can search the course, compare versions and answer the instructor's questions. The access tokens these providers give us are encrypted the same way as the Canvas token.

The assistant and student privacy

Before any text from the course reaches an AI provider, the Workspace replaces each student's name, Canvas login, school ID number and Canvas ID with a stable label such as “Student 14,” and removes email addresses, phone numbers, government ID numbers and links to a student's Canvas profile. The label is made with a secret key that is different for every workspace, and only the instructor's own screen turns the labels back into names. What students wrote (submission text, attachments and comments) is held back from the AI provider unless the instructor turns on reading student work for that course, which also happens when the instructor asks the Workspace to grade that course's work. Names and identifiers stay replaced either way, and the instructor can turn reading student work off in the Workspace's Settings. Scores and submission status may be sent under the label so the assistant can answer questions such as which students are missing work.

This protection has limits. A name that is not on the course's roster, for example a name typed into a document, may not be caught. Pictures, including pictures of student work, are sent to the AI provider as they are, because a name or a face in an image cannot be removed the way text can; the Workspace's Settings says so beside the setting for student work.

Disconnecting and deleting

• An instructor can disconnect Canvas or any file source in the Workspace at any time. After that the Workspace stops reading it, and its files leave the Workspace's file list. The encrypted credential is kept, unusable, as part of our security records until the instructor asks us to delete the workspace.
• To remove Prepara's access on the provider's side as well, the instructor can remove the app in that provider's own settings: for Google, at myaccount.google.com/permissions; for Dropbox, at dropbox.com/account/connected_apps; for Microsoft and Box, in the connected-apps section of that account's settings; for Canvas, by deleting the access token in Canvas under Account, then Settings.
• To delete what the Workspace has stored, including copied files, extracted text, conversations and the course records it read, the instructor emails admin@preparacourse.com. We delete it within 30 days, except the records of who did what and when that we keep for security and legal reasons; after deletion those records no longer point to the instructor's name or email. Copies inside our backups are kept only for disaster recovery and are removed when those backups are deleted.

4. Google API Services and other connected file sources

Prepara's use and transfer of information received from Google APIs to any other app will adhere to the Google API Services User Data Policy, including the Limited Use requirements.

This applies to the files an instructor picks in Google Drive for the Course Workspace and to the name, email address and profile picture Google sends us when someone uses “Sign in with Google.” We hold the same promises for data from Dropbox, OneDrive and Box:

• We use data from Google Drive, Dropbox, OneDrive and Box only to provide the features the instructor asked for, such as searching their course files, comparing versions, answering their questions about the files, and preparing the changes they approve.
• We never sell this data.
• We never use it for advertising, including targeted or personalized advertising.
• We never use it to train, develop or improve general-purpose AI or machine-learning models, whether our own or anyone else's.
• We never use it for research, including the research and model building described in section 7. Those uses do not apply to data from connected file sources.
• We transfer it to others only as needed to provide those features (for example, sending the relevant text to an AI provider, after student names are replaced, so the assistant can answer the instructor's question), for security, to comply with the law, or as part of a merger or sale of our business with the protections described in section 9.
• People at Prepara read this data only with the instructor's permission (for example, to help with a support request the instructor sends us), when it is needed for security, such as investigating abuse, or when the law requires it.

5. AI features

The Service includes AI features that process text or audio you provide, and each one is labeled as AI so you know when you are working with an automated system. AI output can be wrong or incomplete, so treat it as a study aid and check anything important.

AI in the courseware

• The AI tutor. Ordinary questions to the tutor run on Cloudflare Workers AI, which processes your input inside our hosting provider's network using open-weight models. The tutor's longer discussion conversations are answered first by an OpenAI model, with Cloudflare Workers AI as the fallback. A course can also be set to use an Anthropic (Claude) model for its tutor.
• The AI support assistant runs on Cloudflare Workers AI.
• Voice notes are turned into text on Cloudflare's network rather than sent to an outside transcription company.
• Reading aloud. When the tutor reads a reply aloud, the speech is made on Cloudflare Workers AI by default. Where it has been turned on, ElevenLabs makes a higher-quality voice instead.
• In-chapter learning widgets are served by Anthropic (Claude) through a separate proxy. In some books these AI moments are pre-recorded and send nothing to any live model.
• Safety checks. Messages to AI features may pass an automatic check for harmful content, which runs on Cloudflare Workers AI.

When you ask the tutor a question, we also sort it into a general category (for example, asking for help to do the work yourself, or asking the tutor to do it for you) and note which course topics it touches, so we can measure how a class uses the tutor and which topics students find hard. These are short labels; the text of your question is not copied into our learning-activity records. Your tutor conversations themselves are stored with your account and, for a section you are enrolled in, are visible to your instructor along with the rest of your coursework.

AI in the Evaluation Studio and the simulations

The Evaluation Studio turns recordings into transcripts and can give AI feedback on them. Depending on the course, transcripts are made by ElevenLabs, OpenAI or AssemblyAI, and feedback is written by models from OpenAI, Cloudflare Workers AI, Groq, DeepInfra or Anthropic. Some video analysis runs on computers we rent from RunPod and NVIDIA. The simulations use AI characters you talk with, which run on models reached through OpenAI, OpenRouter, Anthropic or Cloudflare Workers AI.

AI in the Course Workspace

The Workspace's assistant runs on OpenAI models. Its question-answering feature sends requests through OpenRouter, a service that passes each request to one of these model makers and returns the answer: OpenAI, Anthropic, Google, and DeepSeek. DeepSeek is a company based in China; it is used only for short, mechanical tasks. As section 3 explains, student names and identifiers are replaced before anything reaches these providers.

Training and outside AI companies

We use each AI provider through its business service for developers, which processes your input to return an answer. We have not given any AI provider permission to use your input, your course material or student data to train its own models. OpenAI, Anthropic and Cloudflare state in their business terms that they do not train their models on data sent through these services. Each provider may keep a request for a limited time under its own terms, for example to detect abuse. We do use AI interactions ourselves, with names and contact details removed, to improve our AI features and to build and test our own models, as section 7 describes, except that data from connected file sources is never used this way.

6. How we use information

We use the information described above to:

• Provide the Service: create and secure your account, sign you in, grant and check your course access, deliver your books, lessons and tools, save your work and your place, and connect your section to your school's LMS.
• Personalize your learning: suggest what to study next, choose practice and review material that fits how you are doing, and remember your preferences.
• Give students and instructors learning analytics: show you your progress, and show your instructor the progress, results and activity of the students in their section, so they can support the class.
• Process payments and keep the order records the law requires.
• Communicate with you: send sign-in, password-reset, receipt and course emails, answer your support requests, and send notifications you have turned on. We do not send marketing email.
• Keep the Service safe: prevent fraud, abuse and cheating, protect course content from unauthorized copying, investigate security problems, and keep audit logs.
• Understand learning behavior and improve the Service: measure what teaches well and change what does not, run the comparisons described in section 7, fix problems, and design new books, lessons and features.
• Build and train our own models and algorithms, for example to recommend study material, to find the topics a class struggles with, to write better practice questions, or to improve the tutor, as section 7 describes.
• Do research on teaching and learning, within the limits in section 7.
• Meet legal duties: comply with the law, respond to lawful requests, enforce our Terms, and protect the rights and safety of our users, our schools and Prepara.

We do not sell personal information, we do not use it for targeted advertising, and we do not build a profile of any student for any purpose other than education.

7. Learning analytics, research and our own models

How we measure learning

To show your progress, to help your instructor support the class, and to measure whether the course works, we record how you read and practice: which chapters, lessons and videos you open, how far you scroll or watch, how much active time you spend, which learning widgets you use, and how you answer checkpoint questions. This record is tied to a pseudonymous ID and stores counts, timings and short labels. It never stores the text of your notes, your answers or your tutor messages. This measurement applies to every account, because the Service needs it to run and to improve.

Your learning profile. From this record we keep a learning profile for each student in each course, such as which kinds of explanation and practice seem to help you most, so the Service can personalize what it suggests. It is used only for your education in the Service.

Program evaluation and experiments. We use this record to measure what teaches well and to change what does not. As part of it, we may assign students to different versions of a lesson, a layout or a study feature and compare how each version works, using an automatic and consistent method that does not depend on who you are. A version may add material or present it differently, but it never removes material you were assigned, so every student always keeps the full educational content of the course. This is evaluation of the product, not research on you as a named individual, and we can end any comparison and return every student to the standard version.

Coarse location. So we can see how a course is used across different regions, we record a coarse, county-level location with your learning activity. We work it out from your general connection details at the moment you read, keep only the county, the wider region and a coarse country, and never store your street address or your exact location. The connection detail we use for this lookup is used for it and then discarded. When we report by place, any county, region or school with fewer than ten students is shown as “fewer than 10” so no small group can be singled out.

Two copies of your learning activity. We keep this record in two forms. A recent working copy in our main database shows your progress and your instructor's class view, and it is deleted automatically after about 180 days (your summarized progress is kept so you do not lose your place). A long-term evaluation copy, which we call the event lake, is a pseudonymous record kept for program evaluation, research and model building; it holds only counts, timings and short labels, never the text of your work and no demographic data. We do not delete the evaluation copy on a schedule, but you can ask us to remove your records from it at any time, as section 14 describes.

Research on teaching and learning and our own models

We study how people learn with Prepara, and we use what we learn to build better books, better features, and our own models and algorithms. For this work we use learning activity (such as reading, answers, time spent, progress and how you use features), AI interactions, survey answers and similar records from across the Service, following these rules:

• We remove identity wherever we can. For research and model building we use pseudonymous, de-identified or aggregated data wherever that is enough for the purpose. De-identified data has had names, email addresses, and other details that could reasonably identify a person removed or replaced. Aggregated data describes groups, not individuals.
• We do not try to re-identify anyone, and we require anyone we share de-identified data with to promise in writing not to try either.
• Identifiable student data is used for research only with the approvals the law or the school requires, such as approval from an institutional review board and the student's consent where it is needed.
• Published research uses de-identified data by default, from October 15, 2026. De-identified learning data may be included in published education research unless you turn this off, as the research choice below describes.
• Research that identifies anyone needs consent. We never publish anything that identifies a student, and we use identifiable student data for research only with the student's consent and any other approval the law or the school requires.
• Aggregated statistics about the Service as a whole (for example, the average time students spend on a chapter across all of our books, or how often a kind of question is answered correctly) cannot identify anyone. We may publish or share them, for example with schools, in reports, or in descriptions of our products.
• Data from connected file sources is never used for any of this work, as section 4 explains.

We own the de-identified and aggregated information we create, and we may use it for any lawful purpose, subject to your research choice and to the promises in section 4. If you send us feedback or suggestions about the Service, we may use them freely and without any obligation to you.

Your research choice

Separately from the measurement above, which applies to every account so the Service can run and improve, you choose whether your de-identified learning data is included in published education research. From October 15, 2026, this setting is on unless you turn it off. You can turn it off at any time in Settings, and from then on your records are left out of research going forward. Published results report groups, never individuals, and your choice does not affect your access or your grade.

Two limits apply to this change:

• Earlier activity keeps the earlier rule. Learning activity recorded before October 15, 2026 follows the choice as it stood when it was recorded: it is left out of research unless you had turned research inclusion on.
• A choice you already made stays. If you turned research inclusion off before this change, including by answering no on the question we used to show at first sign-in, it stays off until you turn it on in Settings. If you turned it on, it stays on.

We no longer ask about research on a card when you first sign in. The setting is in Settings, under your data.

Surveys and confidence check-ins

From time to time the Service asks short, optional questions inside the reader: a brief confidence or AI-literacy survey at the start and end of a course, and one-question check-ins after some chapters. These are voluntary and you can skip them, each is one screen or one question, and each states what it is for. Your answers are rating-scale and short responses, kept as part of your record so your school can use them as learning evidence and so we can improve the course. We label each questionnaire with a version, so if we reword one, older answers are never mixed up with new ones.

8. Cookies, browser storage and website traffic

Cookies. We use only cookies the Service needs to work: a sign-in cookie that keeps you logged in, a separate sign-in cookie used when Prepara opens inside your LMS, and a technical cookie that makes sure you see your own latest changes. We do not use advertising cookies or cookies that track you across other websites.

Browser storage. The app keeps some information in your browser's own storage: your display preferences, your place in what you were reading, and a random install identifier created the first time you use the app on that browser. The install identifier is not your name or your account; it tells us which of your devices an activity came from, so your learning activity and device statistics are counted correctly. You can clear it by clearing your browser's site data, and the app will create a new one.

How we measure website traffic. To understand how many people visit our public pages and where they come from, we use a method that counts visits without identifying people, and we store nothing that could re-identify you. We do not use cookies for this, and nothing is stored in your browser. For each visit we create a one-way value from a secret that changes every day plus general request details (the page, a coarse country and the device type). Because that daily secret is deleted, the same visitor becomes a new, unlinkable value the next day, and we never store your IP address or your full browser details for this purpose. Because this method uses no cookies and cannot follow you across days or across other websites, it does not need a cookie consent banner.

Bot protection. Some forms, such as the sign-in form, can use Cloudflare Turnstile to check that a person, not an automated program, is filling them in.

Content your instructor adds. An instructor can show material from other services inside a class activity, such as a YouTube or Vimeo video, a Google document, or a PowerPoint file. That service's own privacy policy applies to what it collects when you view it. We load YouTube videos from YouTube's reduced-tracking address.

9. Who can see your information

• You can see all of your own account details, notes, highlights, voice notes, boards and activity.
• Your instructor and teaching staff for a section you are enrolled in can see your coursework and learning activity for that section: your progress, your Checkpoint Quiz and assignment results, your Checkpoint Quiz answers, your submissions, and your AI tutor conversations. They cannot see the private study material you keep for yourself, such as your highlights, personal notes, voice notes and canvas boards. They see answers to the confidence questions only as class totals.
• Your class. In an in-class activity, your instructor sees your answers, and answers may be shown to the class on the shared screen, for example in a word cloud or a results chart. Ratings and comments you give in a peer evaluation go to your instructor, and your instructor decides what, if anything, the students you rated can see. In the Evaluation Studio, your recordings are seen by your instructor and by anyone your instructor assigns to review them.
• Your school can receive records about its own students and classes when it directs us to, for its own purposes, such as grades sent to its LMS or reports it asks for.
• Our staff see personal information only when they need it to run, support or secure the Service, and they are bound by confidentiality.
• Service providers listed in section 10 process information only to do their job for us.

Beyond these cases, we share personal information only:

• when you ask us to or give your consent;
• when your school directs us to, for its own records;
• when the law requires it, such as a valid subpoena or court order, or when we believe in good faith that sharing is needed to protect someone's safety, to investigate fraud or security problems, or to protect the rights of our users or of Prepara;
• if Prepara is involved in a merger, acquisition, financing or sale of all or part of its business, in which case the information may be transferred to the new owner, which must keep the promises in this policy for information collected under it, including the student-data promises in section 11; and
• as de-identified or aggregated information that cannot reasonably identify you, as section 7 describes.

We never sell personal information, and we never share it for targeted or cross-site behavioral advertising.

10. Service providers we use

We use a small set of providers to run the Service. Each one processes only the information it needs for its job, under confidentiality and security obligations, and we never sell personal information to any of them.

• Cloudflare (United States, with a worldwide network): hosting and computing for every part of the Service, our databases, file storage and queues, the event lake, the anonymous traffic count, bot protection, the live video rooms in the Evaluation Studio, and Cloudflare Workers AI, which runs the AI tutor's default models, the support assistant, voice-note transcription, reading aloud and safety checks.
• OpenAI (United States): the AI tutor's discussion conversations, the Course Workspace's assistant, and transcription, feedback and AI characters in the Evaluation Studio and the simulations.
• Anthropic (United States): the in-chapter learning widgets, the AI tutor where a course uses a Claude model, and some AI work in the Evaluation Studio, the Selling Simulation and the Course Workspace.
• OpenRouter (United States): passes AI requests from the Course Workspace and the simulations to the model maker being used, including Google and DeepSeek for the Course Workspace.
• Groq, DeepInfra, AssemblyAI, RunPod and NVIDIA (United States): AI feedback, transcription and video analysis in the Evaluation Studio.
• ElevenLabs: higher-quality reading aloud where it has been turned on, and transcription in the Evaluation Studio and the Selling Simulation. When it is not turned on for reading aloud, text-to-speech runs on Cloudflare Workers AI and nothing is sent to ElevenLabs.
• Zoom (United States): an optional way to run live rooms in the Evaluation Studio.
• Neon (United States): the database for the simulations.
• Stripe (United States): payment processing on behalf of PhD Analytics LLC. Stripe handles card data directly through a redirect checkout (PCI DSS SAQ-A); your card number never reaches our servers.
• Postmark, with Resend as a backup (United States): delivery of transactional email only, such as sign-in links, password resets, receipts, invitations and course notifications. We do not send marketing email.
• Backblaze B2 (United States): encrypted, off-site copies of our database and of the pseudonymous learning-activity records, kept for disaster recovery.
• Hetzner: a server we rent that relays grade transfers from Prepara to your school's LMS.
• Browser push services (such as those run by Apple, Google and Mozilla): deliver the browser notifications you have turned on. Your browser chooses the service.

Your school's LMS, such as Canvas, and the file services an instructor connects to the Course Workspace are not our service providers. They are systems your school or the instructor chose, and their own terms apply to them.

11. Student records, FERPA and state student privacy laws

When your school uses the Service for a class, the records we keep about you in that class (such as your enrollment, your progress, your checkpoint scores, your assignment submissions and your tutor conversations for that section) can be education records under FERPA, the federal student-privacy law. For those records we act as a school official with a legitimate educational interest, under the school's direct control for the use and upkeep of those records. That means:

• We use education records only to provide and improve the Service for the school and for the other purposes FERPA allows, including creating de-identified data, which is no longer an education record once students cannot reasonably be identified from it.
• We do not share education records except as the school directs, as FERPA allows, or as the law requires.
• Your instructor and teaching staff for a section can see the coursework and learning activity of the students enrolled in that section, the same way they would in any class.
• The school keeps control of its students' records. A school may ask us to provide, correct or delete a student's records, and we will do so. If you want to see or correct your education records, you can ask your school, and we will help your school answer.
• We will sign a data processing agreement with a school on request.
• When an instructor uses the Course Workspace, the student information it reads from that instructor's Canvas course is used only to provide the Workspace to that instructor. We do not use identifiable student information from an instructor's Canvas course for research or to train models.

Our student privacy commitments. Many states have student privacy laws, such as California's Student Online Personal Information Protection Act. Wherever you live, we make these commitments about student information:

• We do not use student information for targeted advertising.
• We do not sell student information.
• We do not build a profile of a student for any purpose other than education.
• We do not share student information except as this policy describes, for the school's purposes, or as the law requires.
• We protect student information with reasonable security, as section 13 describes, and we delete it when the school asks, as section 14 describes.

12. Learning-management systems (LTI)

If your instructor connects a course section to your school's LMS, such as Canvas, using the LTI 1.3 standard, we exchange data with that LMS to make the connection work. When you launch the Service from your LMS, your LMS identity (such as your name, email and role in the course) is shared with us so we can sign you in to the right section. We may read the course roster and, where the instructor allows it, the course's groups, so the section in Prepara matches the class. Your grades and completion for that section go only to your school's own gradebook, through a relay server described in section 10. This happens only for sections an instructor has connected and only with your own school's LMS. An instructor may also give Prepara an optional Canvas access token so it can read the course's groups; we encrypt that token and never show it on any page. Your school's own privacy policy governs the data once it is in your LMS.

13. Security

• Data is encrypted in transit and at rest.
• Passwords are stored only as secure hashes. Two-factor authentication and passkeys are available.
• Credentials we hold for other services, such as Canvas access tokens and the tokens for connected file sources, are encrypted with AES-256-GCM, each under its own key, and those keys are encrypted by a master key that exists only as a server secret.
• Course content is available only to signed-in users who have access, and each copy shows a visible watermark and an invisible watermark tied to the user.
• Our staff get the least access they need, and we keep an audit log of access and administrative actions.

To report a security issue, email admin@preparacourse.com; our security contact and disclosure policy are also published at /.well-known/security.txt. No system is perfectly secure, but if a breach affecting your personal data happens, we will tell you and any required authorities without undue delay and as the law requires. Where a school's data is involved, we will notify the school promptly, within 72 hours of becoming aware, so it can meet its own obligations.

14. How long we keep information

We keep your account details for as long as your account is active. For other kinds of data we keep only what we still need:

• Support conversations (your messages with our support team and the AI support assistant) are kept for about 180 days and then deleted automatically.
• The recent copy of your reading activity in our main database is kept for about 180 days and then deleted automatically. Your summarized progress is kept so you do not lose your place in a course.
• The long-term evaluation copy of reading activity (the event lake) is kept for program evaluation, research and model building. It holds only pseudonymous counts, timings and short labels, never your name beside them and never the text of your work.
• Your notes, highlights, voice notes and transcripts, and boards are kept while your account is active. You can delete any of them yourself at any time.
• Coursework, results and tutor conversations for a class are kept while your account is active, so you and your instructor can see them, unless your school asks us to delete them sooner.
• Survey and confidence-check answers are kept as part of your record while your account is active.
• Recordings in the Evaluation Studio and the simulations are kept while the course that uses them needs them, and are deleted with your account or when you, your instructor or your school asks.
• Course Workspace data (the instructor's connections, course records, copied files and conversations) is kept while the instructor's workspace is active and deleted as section 3 describes.
• Order and payment records are kept as long as the law requires for accounting and tax.
• Security and audit logs are kept for up to about three years.
• Backups are kept only for disaster recovery, and deleted information leaves them as those backups are replaced.

When you delete your account, we remove your personal data, your notes, highlights, voice notes, boards, and the activity records tied to your account in our main database within 30 days, except the order, payment, security and audit records we must keep for legal reasons. While your account is part of a class whose term is still running, deletion waits until the term ends, as the Terms of Service explain; you can still download your data at any time. Deleting your account also stops any further use of your records for research. The long-term evaluation copy holds only pseudonymous counts and labels and never the text of your work; we do not delete it automatically, but if you want it removed, email admin@preparacourse.com and we will delete your records from it. De-identified and aggregated information that no longer identifies you may be kept. Records we hold on behalf of a school are returned or deleted on the school's instruction, or within 30 days after that school's contract with us ends.

15. Your privacy rights and choices

These rights apply to everyone, wherever you live:

• Access and download: use Download my data in Settings for a full export, or ask us for a copy of the personal information we hold about you.
• Correct: update your profile and password in Settings, or ask us to correct anything else.
• Delete: delete your account in Settings (subject to the class-term timing in section 14), delete your own notes and other study material at any time, ask us to remove your records from the long-term evaluation copy, or ask us to delete other information.
• Research choice: turn research inclusion on or off in Settings at any time.
• Notifications: turn browser notifications off at any time in your browser's settings for our site.
• Connected accounts: instructors can disconnect Canvas or any file source from the Course Workspace, and remove Prepara's access in that provider's own settings, as section 3 describes.
• No penalty: we will not treat you differently, in price, access or grades, for using any of these rights.

Where you use the Service through your school, that school relationship and FERPA govern your class records, and some requests, such as deleting records of a class, may need to go through your school. We will tell you if that is the case and help you reach the right person.

California residents

If you live in California, the California Consumer Privacy Act, as amended by the California Privacy Rights Act, gives you these rights, in addition to those above:

• The right to know what personal information we have collected about you, where it came from, why we use it, and whom we disclose it to, and to get a copy of it.
• The right to delete personal information we collected from you, with the exceptions the law allows.
• The right to correct inaccurate personal information.
• The right to opt out of the sale or sharing of personal information. We do not sell personal information, and we do not share it for cross-context behavioral advertising. We have not done either in the past 12 months, and we have no actual knowledge of selling or sharing the personal information of anyone under 16.
• The right to limit the use of sensitive personal information. The only sensitive personal information we collect is your account login (your email address with your password or passkey). We use it only to sign you in and keep your account secure, which the law permits without an opt-out.
• The right not to be discriminated against for using these rights.

In the past 12 months we collected these categories of personal information, from you, from your school and its LMS, and from your use of the Service, for the purposes in section 6: identifiers (such as name, email address, account ID, LMS ID and IP address); education information (such as enrollment, coursework, results and learning activity); internet and device activity (such as pages and features used, and device type); audio and video (voice notes and recordings); general geolocation (county level only); commercial information (order records); professional information for instructors (school, courses and role); and inferences used to personalize learning (the learning profile in section 7). We disclosed these categories for business purposes to the service providers in section 10 and, for class records, to your school and instructor.

You may use an authorized agent to make a request for you. We will verify your identity (and the agent's authority) by confirming your control of the email address on your account before we answer, and we will respond within 45 days, or tell you if we need more time as the law allows.

Other U.S. states

Residents of other states with consumer privacy laws, such as Colorado, Connecticut, Virginia and Texas, have similar rights to access, correct, delete and get a copy of their personal information, and to opt out of sale, targeted advertising and certain profiling. We honor these requests for everyone. We do not sell personal information and do not use it for targeted advertising. Checkpoint and quiz scores are calculated automatically from your answers, and your instructor decides how they count toward your grade. If we turn down your request, you may appeal by replying to our answer, and we will review it and tell you the outcome.

Visitors from the European Economic Area, the United Kingdom and Switzerland

Prepara is designed for colleges and universities in the United States. If data protection law in the European Economic Area, the United Kingdom or Switzerland applies to your information, Prepara LLC is the controller, except where we process information on behalf of your school, in which case your school is the controller. We process your information to perform our agreement with you (providing the Service), for our legitimate interests (keeping the Service secure, understanding how it is used, and improving it, balanced against your rights), and to meet legal obligations. Published research uses only de-identified data, which is no longer personal data once no one can reasonably be identified from it; research that identifies anyone uses personal data only with that person's consent. You have the right to access, correct, delete, restrict or object to our processing of your information, to receive it in a portable format, and to withdraw consent at any time without affecting earlier processing. You may also complain to your local data protection authority.

How to use your rights

Most of these you can do yourself in Settings. For anything else, email admin@preparacourse.com from the address on your account, tell us what you would like, and we will honor your rights as the law requires.

16. Where your information is processed

Prepara is based in the United States, and we store and process information in the United States and on our hosting provider's worldwide network. Some AI requests from the Course Workspace, with student names and identifiers already replaced, may be processed by DeepSeek in China, as section 5 explains. If you use the Service from outside the United States, your information will be transferred to the United States, where privacy laws may differ from those where you live. Where the law requires it, we use appropriate safeguards for these transfers.

17. Younger users

You must be at least 13 years old to have an account. The Service is built for college and university coursework and is not directed to children under 13, and we do not knowingly collect personal information from children under 13. Children's privacy laws such as COPPA apply to children under 13 and do not apply to college students, but we still hold to a 13-and-over rule and will delete any account we learn belongs to a younger child. If you believe a child under 13 has given us personal information, email admin@preparacourse.com and we will delete it.

18. Changes to this policy

We may update this policy as the Service changes or the law changes. When we make a change, we update the dates at the top of this page. When we make a material change, we will also give notice through the Service or by email before it takes effect. The first time you sign in, and again after a material change, we ask you to confirm that you have read and agree to the Terms of Service and this Privacy Policy before you continue, and we keep a record of that agreement. Using the Service after a change takes effect means you accept it. We will not use personal information we collected under an earlier version of this policy in a materially different way without the notice and, where the law or a school agreement requires it, the consent that this calls for.

19. How to contact us

Questions, requests or complaints about this policy or your information go to admin@preparacourse.com. We are Prepara LLC, an Oklahoma company. Schools that want a data processing agreement, or IT staff reviewing Prepara for their institution, can reach us at the same address.