Managing Files

Transform your digital twin into a personalized knowledge expert by uploading files to your secure IP Vault. Your digital twin can then reference this content to provide informed, contextual responses tailored to your organization and needs.

Your IP Vault provides enterprise-grade security, ensuring your proprietary information remains completely isolated to your digital twin instance.

File Access Control

Effective March 11, 2026 — In an effort to continuously secure the IP Vault, all files are now protected by a three-tier access control system.

Every file in your IP Vault is assigned one of three access levels:

Access Level	Who Can View	Description
Public	Anyone with the link	The file owner has explicitly opted to make the file publicly accessible. Anyone with the URL can view it.
Private (default)	Signed-in members	Requires sign-in and an active membership to the digital twin associated with the file. This is the default for all files.
Confidential	File owner only	Restricted to the file owner. No other users — including administrators — can access it.

Files must be explicitly set to Public if you want to share their URL with third parties. Accessing a file that is not set to public without being signed in will display a Sign-in Required page.

You can change a file’s access level from its file actions menu. To change the access level for all files in a collection at once, use the Make All Public or Make All Private options in the collection menu.

Files accessed through the IP Vault will load in a preview when the content type is supported.

If you attempt to access a file without the required permissions, you will see a denial page explaining why access was blocked and how to resolve it.

---

Vault Tabs

Your IP Vault organizes files into three tabs based on scope:

Tab	Visibility	Who Can Upload
My Files	Only you	You
Institution Files	All users in your institution	Administrators
Account Files	All users across all institutions in your account	Account administrators

Your Digital Twin automatically references files from all three tabs during conversations. Institution and Account files allow administrators to share common knowledge resources — such as policies, handbooks, or course materials — with all users.

Vault Health Indicators

Each tab displays a health indicator showing your file counts in the format included / total:

• Included — Files actively used by your Digital Twin for RAG responses
• Total — All files in that vault (excludes deleted items)

For example, My Files - 3 / 5 means 3 of your 5 personal files are currently included in your knowledge base. Toggle individual files on or off using the Exclude from RAG action in File Actions.

---

Adding Files to Your IP Vault

1. Access Your File Manager

   Click the Files button on the text input bar to open your IP Vault.

   
2. Start Adding Content

   In your IP Vault dashboard, click the + Add button at the bottom of the page.

   
3. Choose Your Upload Method

   

   Choose from Device, URL, or Clipboard upload options. See File Upload Types below for detailed instructions on each method.

4. Complete the Upload

   After selecting your files, click Upload to begin the process.

   

   A live progress indicator tracks each file through its processing stages. File ingestion may take several minutes depending on file size and type.

   

   Each file progresses through these stages:

   1. Uploading — File is being transferred to the server
   2. Processing — File content is being extracted and indexed for RAG
   3. Included — File is ready and available for your Digital Twin to reference
   4. Failed — Processing encountered an error (check the file format and try again)

   When complete, your file status will display as “Success” and be ready for use. Click Done in the lower-left corner to go back to your IP Vault.

   
5. Verify Files in Your Vault

   Your uploaded files now appear in your IP Vault, ready for your digital twin to reference.

   

Supported File Formats

Your IP Vault accepts a comprehensive range of file types:

Content Type	Supported Formats
Images	PNG, JPG, JPEG, WEBP, GIF, BMP, TIFF, SVG, ICO
Audio	MP3, M4A, WAV, WEBM, OGG, FLAC, AAC, MPGA
Video	MP4, WEBM, MPEG, MOV, AVI, MKV, WMV, OGV
Documents	PDF*, DOC, DOCX, ODT, RTF, Pages
Spreadsheets	XLS, XLSX, ODS, CSV, TSV, Numbers
Presentations	PPT, PPTX, ODP, Keynote
Email	EML, MSG, MHT, MHTML, MIME
Web Content	HTML, HTM, XHTML, CSS, RSS
Text & Data	TXT, MD, JSON, XML, YAML, VTT
Code	IPYNB, JS, JSX, TS, PY, C, CPP
Learning Content	IMSCC (IMS Common Cartridge)
Archives	ZIP, GZ, 7Z, RAR, TAR, BZ2
URL Sources	Web pages, Google Docs/Sheets/Slides, YouTube videos

Warning: PDF files containing no text, or only images (image-based PDFs) will be processed through a process called Optical Character Recognition a.k.a OCR.

Maximum upload limit: 500 MB total across all files

Audio & Video Processing: Audio and video files of any size are automatically transcribed and chunked into separate text segments for RAG indexing. Currently, Pria extracts speech-to-text transcriptions from media files. Advanced features — including cue point detection, speaker diarization, and timestamp linking — are in active development.

YouTube Videos: When you paste a YouTube URL (youtube.com/watch, youtu.be, or youtube.com/shorts), Pria automatically extracts the video transcript with [M:SS] timestamps, video title, channel name, and thumbnail. The transcript is stored as a text file that your Digital Twin can search and reference. If the video has captions disabled, you can still import it with Skip Indexing to store the metadata without generating embeddings.

iPhone & QuickTime Recordings: .mov files and other container formats that aren’t natively supported by transcription providers are automatically converted to a compatible audio format before processing. This means you can upload voice memos recorded on your iPhone or screen recordings from QuickTime and they will be transcribed just like any other audio file.

File Upload Types

Your IP Vault supports three convenient methods for adding content, each optimized for different use cases:

Device Upload

Upload files directly from your computer:

1. Click Device in the upload dialogue
2. Browse and select files from your local storage
3. Click Upload to add them to your vault

Ideal for documents, presentations, and media files stored locally.

Import Web Content via URL

Easily import content from any web page or YouTube video:

1. In the upload dialogue, select the URL tab
2. Enter the complete web page address or YouTube video URL
3. Select Scrape site to include all dependent pages (for web pages)
4. Click Import to add the content to your vault

Google Docs, Sheets & Slides — Paste a Google document URL and Pria will extract the full content. If you have connected your Google account via Settings > Connected Apps, Pria uses your Google authentication to access private and shared documents. Without a connected account, only publicly shared documents can be imported.

YouTube URLs are automatically detected — Pria extracts the video transcript with timestamps, metadata (title, channel), and thumbnail. If the video has no captions available, you can still import it with Skip Indexing enabled to store the metadata without embeddings.

Perfect for articles, documentation, online resources, Google documents, and video transcripts.

Paste from Clipboard

For quick content addition:

1. Copy any text content to your clipboard
2. Select the Clipboard option in the upload dialogue
3. Paste and save your content

Great for quick snippets, notes, and copied text.

These features are perfect for importing articles, documentation, or any web-based content your digital twin should reference.

Verify scraped content: After uploading via URL, review the extracted text content. If the scraped content appears as garbage (containing only JavaScript instructions), the source page may be protected by a content management system. In this case, export the content directly from the source site and upload it as a file instead.

Remember that the quality of the documents you put-in determins the quality of the response that come-out.

Managing Your Files

Each file in your IP Vault comes with powerful management options:

File Actions Explained

• Exclude from RAG - Toggle whether your Digital Twin can reference this file in responses
• Make Public - Set the file to Public access so anyone with the link can view it
• Make Private - Set the file to Private access so only signed-in members can view it
• Rename - Update the file name for better organization
• Download - Save a copy to your device
• Delete - Permanently remove the file from your vault
• Reprocess Content - Re-extract and re-index the file’s content for RAG. Opens a confirmation dialog with options to reprocess from existing text or re-ingest from URL when available
• Move to Instance - Transfer the file to an institution’s shared file collection (admin only)
• Copy Name - Copy the file name to clipboard
• Copy URL - Copy the file’s shareable URL to clipboard. The recipient must have appropriate access to view the file
• RAG Segments - View the text chunks extracted from this file for RAG indexing
• View Details - See comprehensive file information including size, type, upload date, and processing status

RAG Segments Panel

The RAG Segments panel lets you view and manage the text chunks that your Digital Twin uses for retrieval. Open it from the RAG Segments action in any file’s context menu.

Editing Title and Summary

At the top of the panel you can edit the file’s Title and Summary — these metadata fields help your Digital Twin understand the file’s context. Changes are saved when you click Update in the footer.

Tokens Used

A token counter badge in the panel header shows the total tokens consumed by processing operations on this file (embedding generation, sanitization, etc.). The counter updates live as you sanitize segments.

Sanitize Segments

Each RAG segment has a sanitize button (sparkle icon) that uses AI to clean up the segment text. This is useful for segments that contain noise from web scraping, PDF extraction, or document conversion — such as navigation menus, cookie banners, broken formatting, or encoding artifacts.

How to use:

1. Open the RAG Segments panel for any file
2. Click the sparkle icon next to a segment’s character counter
3. The AI will clean the text while preserving all substantive content
4. Review the cleaned text in the editor
5. Click Update to save the changes, or navigate away to discard

Sanitization does not summarize or shorten your content — it only removes noise and fixes formatting. Token usage for the AI call is tracked on the file’s token counter.

Sanitize All

To clean up an entire file at once, click the Sanitize All button in the panel footer. This runs AI sanitization on all segments concurrently, skipping any segments that you have already edited manually. A spinner indicates progress, and the token counter updates as each segment completes. Each segment can be individually undone after sanitization.

Unsaved Changes Protection

If you attempt to close the RAG Segments panel with unsaved changes (edited segments, title, or summary), a confirmation dialog will ask whether to Keep Editing or Discard & Close — preventing accidental loss of work.

Reprocess Content

The Reprocess Content action in a file’s context menu opens a dialog with two options:

• Re-embed — Uses the original extracted text to regenerate embeddings and metadata. This overwrites any manual customizations you made to segments.
• Re-ingest from URL — Available only for files originally imported via URL. Re-downloads fresh content from the source URL, replaces the file on disk, and re-runs the full ingestion pipeline. The existing file record is updated in-place (no duplicate is created). Collection membership is preserved. For YouTube URLs, the transcript, thumbnail, and metadata are refreshed from the source.

A Skip Indexing checkbox is available on the Re-ingest option. When enabled, the file content is re-downloaded but embedding generation is skipped — useful when you want to refresh the stored file without updating the knowledge base, or when re-ingesting a YouTube video that has no captions available.

Re-ingesting from URL is useful when the source content has been updated and you want your Digital Twin’s knowledge base to reflect the latest version. If re-ingestion fails (e.g., the source URL is unavailable), the file’s original status is automatically restored.

Sorting and View Options

Your IP Vault includes sorting and display controls to help you organize and browse files efficiently:

Control	Function	Options
Sort by Date	Order files by upload date	Newest first (↓) or oldest first (↑)
Sort by Name	Order files alphabetically	A → Z (↑) or Z → A (↓)
Grid/Compact View	Toggle display density	Grid view or compact list

Click the active sort button again to reverse the order (toggle between ascending and descending).

Collection Highlighting

When browsing collections, folders that contain included files (files actively used for RAG) are visually highlighted with a stronger blue border, making it easy to distinguish active collections from empty ones at a glance.

---

Bulk Operations

Select multiple files to perform operations on them all at once. Use the checkboxes on file cards or the “Select All” option to choose files, then use the bulk action toolbar.

Available bulk actions:

Action	Description
Include	Add selected files to the AI’s knowledge base (RAG indexing)
Exclude	Remove selected files from the AI’s knowledge base without deleting them
Delete	Permanently delete selected files
Move to Collection	Move selected files into a collection (folder)
Make Public	Make selected files accessible via public download link
Make Private	Restrict selected files to authenticated access only
Reprocess	Re-index selected files in the AI knowledge base (useful after content updates)

Download as ZIP

You can download an entire collection as a ZIP archive. This includes all files in the collection and any nested sub-collections, preserving the folder structure.

ZIP downloads have a maximum size of 1 GB. For very large collections, download sub-collections individually.

---

Import Course Packages (IMSCC)

Pria supports importing IMS Common Cartridge (.imscc) course packages — the standard export format used by learning management systems like Canvas, Blackboard, Moodle, and D2L.

How it works:

1. Upload a .imscc file to your IP Vault
2. Pria automatically extracts the course content:
   • Syllabus and course overview pages
   • HTML content from lessons, assignments, and modules
   • Images and media embedded in course pages
   • Discussion topics and announcements
3. Extracted content is organized into collections and indexed for AI retrieval

This is a powerful feature for migrating course content from another LMS into Pria. Your Digital Twin can immediately reference the extracted course materials in conversations.

Image-heavy courses may take longer to process, as Pria uses AI vision models to generate descriptions for embedded images.

---

Best Practices

Organize Effectively

Use descriptive file names and detailed descriptions to improve AI-powered search and retrieval

Control Access

Use file access levels — Public, Private, or Confidential — to control who can view your files. Set files to Public only when you intend to share them externally

Build Comprehensively

Add diverse content types to create a rich knowledge base for more intelligent responses

Keep Current

Regularly update and refresh dynamic content to maintain accuracy

Your digital twin becomes more intelligent with every file you add! Build a comprehensive knowledge base that enables quick, accurate, and contextually relevant responses.