Your data / A practical explanation
Your work deserves
a clear explanation.
Here is how this preview handles textbooks, study records, and connected AI services. This page describes the product’s current behavior.
What is stored
The deployment stores your uploaded PDFs, extracted passages, study conversations, notes, practice attempts, learning history, and profile. Account data is separated in PostgreSQL. The deployment operator maintains the server and backups and can administer this storage.
What leaves the deployment
Clerk handles hosted sign-in and identity. When you use AI tutoring, preparation, or search, the relevant content is sent to the configured Anthropic or OpenAI service. Natural AI narration sends the selected page segment to OpenAI. The providers’ own data-handling terms apply; StudeeBuddy cannot promise that a provider retains nothing.
The landing-page sample uses fixed feedback and sends no AI request. The open demo course is free to read. Asking its tutor requires sign-in and sends your question, previous demo conversation and prepared course material to Anthropic, paid for by StudeeBuddy. Demo questions and answers are stored with your account and are included in study-data export. Device read-aloud uses your browser’s speech service, which may be local or network-based depending on the chosen voice.
AI connections
StudeeBuddy manages provider connections using operator-owned credentials. Customers do not submit personal API keys. Historical encrypted keys from an earlier version are no longer used and can be removed through workspace erasure.
Export and removal
Settings lets you export your study records as JSON. Download original PDFs from the reader separately. Hosted accounts can erase their StudeeBuddy workspace after an explicit confirmation. That removes live account study records, demo conversations and saved keys, while leaving the Clerk sign-in identity available. We retain the demo-question count to prevent resetting the free allowance. Platform demo-expense records contain no question text or answers and are retained for budget auditing.
Original PDFs are stored in the configured private S3 bucket, or on the server when S3 is disabled. S3 files are grouped by account under this app’s own prefix; the server also keeps a reading cache. Workspace erasure removes live study records and saved API keys, but does not immediately purge original files or caches. Older copies can remain in database/file backups and S3 versions, depending on the operator’s retention configuration. Contact the deployment operator for physical-file removal and backup-retention questions. Removing an identity through Clerk alone does not erase the StudeeBuddy workspace.
Sign in to manage your dataBrowser drafts and audio
Your browser keeps study drafts for recovery and clears StudeeBuddy drafts when the signed-in identity changes. The sample completion and optional note stay in browser storage. Generated audio stays in the current player’s memory; it is not a permanent audiobook library.
Product analytics
When enabled by the deployment operator, PostHog receives page categories, explicit feature actions, request outcomes and AI usage metadata such as model, token counts, estimated cost and latency. Signed-in activity uses your account identifier. This integration does not record your screen, capture typing, or send document text, drafts, chat content, email addresses or authentication URLs. Browser analytics respects Do Not Track and Global Privacy Control. PostHog receives network requests as a separate service; GeoIP enrichment is disabled.
AI is an aid, not an authority
Check citations and worked answers. You can question grading and inspect learning evidence. StudeeBuddy does not guarantee that every generated explanation, narration, or exercise is correct.
Help and feedback