Simply AskingDocs

Add Content

Three ways to bring knowledge in: Upload what already exists, Create what does not exist yet, or Sync from an app you already use. Watch progress from the Uploads tray on the same page.

The Add page asks one question: where is this content right now? There are three answers, and each is a button at the top of the page.

UploadIt already exists, as a file on your computer or at a web address.
CreateIt does not exist yet. Write it, dictate it, or have Lumen draft it.
Sync from an appIt lives in a tool that keeps changing, like Google Drive or Notion.

After you add something, open the Uploads tray on the Add page to see whether it finished. Uploads is not in the app header; it is a tray on Add.

Q4 Financial Report.pdf2.4 MB
Product Roadmap.docx890 KB
Team Photo.jpg3.1 MB

Drop files to add instantly

PDF, Word, Excel, images, audio. Drag and drop or browse. Batch upload detects document families automatically.

Pro Tip: What Happens After Upload: Your document goes through text extraction, sectioning, embedding, entity extraction, relationship mapping, and graph integration. Entities (people, organizations, concepts, locations) are automatically identified and connected to your knowledge graph. Knowledge gaps are detected automatically as the system identifies areas needing more coverage.

Add your first document in seconds:

1

Open Add page

Click Add in the center action row

2

Pick where the content is now

Click Upload, Create, or Sync from an app

3

Add your content

Follow the prompts in the pane that opens, then upload or publish

Click Upload to open a single pane with two ways in: a drop zone for files from your computer, and a URL field beneath it for content on the web. Use whichever fits. Once you stage a file the URL field steps aside, and once you paste a link the drop zone does, so there is only ever one thing to submit.

Files from your computer

Upload documents, audio files, images, spreadsheets, and more. We support 20+ file types.

DocumentsPDF, DOC, DOCX, TXT, MD, RTF, Pages
SpreadsheetsXLS, XLSX, CSV, Numbers
PresentationsPPT, PPTX, Keynote
ImagesJPG, JPEG, PNG, GIF, BMP, TIFF
AudioMP3, M4A, WAV, OGG, FLAC, AAC, WMA
VideoMP4, MOV, AVI, WebM
1

Add your files

Drag files onto the drop zone, or click it to browse your computer

2

Add details

Add a title (optional, auto-generated from filename), select a document type, and choose an access group if applicable

3

Upload

Click Upload to start processing

Before uploading, click a file's Preview button to check its contents. This works for single files, batches, and files inside a selected folder. Only the file you open is previewed, locally on your device. PDFs, Word documents, spreadsheets, text, images, audio and video use the available document viewer. PowerPoint files show a simplified preview. Some older or specialist formats cannot be previewed, but can still be uploaded.

Pro Tip: Tip: You can upload multiple files at once. Batch upload automatically detects document families (like purchase orders from different vendors) and suggests document types. Maximum file size is 50MB per file.
Important: Image Upload Requirements: Images (JPG, PNG, GIF, WebP, BMP, TIFF, SVG) require a summary description before uploading. Unlike documents with extractable text, images rely on your summary and tags for search. The Summary field appears automatically when you select an image file.

Content from a web address

Below the drop zone is a field marked Paste a URL to import. Drop a link in and the content is fetched, extracted, and summarized for you.

Web PagesArticles, blog posts, documentation. AI summarizes for you.
YouTube VideosTranscribed so you can search and chat with the content.
1

Paste the link

In the Upload pane, paste the web address into the URL field below the drop zone

2

Review details

A title is generated from the URL. Edit it if needed. AI generates a summary automatically.

3

Import

Click Import to fetch and process the content

Important: Note: Some websites block automatic fetching. If a URL fails, try copying the text and writing it up in Create instead.

Click Create to open a full rich text editor. Everything you need to produce a document lives in that one editor: type it yourself, dictate it with your voice, or have Lumen draft a first pass for you to edit.

1

Open Create

Add a title and start writing in the editor

2

Format your content

Use the toolbar for bold, italic, headings, lists, tables, images, and code blocks

3

Enhance with AI (optional)

Click Enhance to fix spelling, simplify, expand, or improve formatting

4

Publish

Click Publish to save and process

Dictating instead of typing

The editor toolbar has a Voice button. Click it and speak, and your words are transcribed straight into the document you are writing. Click it again to stop. The keyboard shortcut is Cmd+Shift+V (Ctrl+Shift+V on Windows). Dictation adds text to your draft, so you can mix talking and typing freely in the same document.

Recording a voice note

When the editor is empty you will also see a Record a note button. This is the other kind of voice: instead of dictating into a document, the recording becomes the document. Press the microphone to start, stop when you are done, then give it a title and save. You get an audio document with its transcript attached, so it is searchable like anything else and the original audio is kept.

Pro Tip: Two voice buttons, two different jobs. The toolbar Voice button adds your words to the document you are writing. Record a note makes the recording itself the document. The second one only appears while the editor is empty, because it is a way to start a document rather than a way to add to one.
Pro Tip: Already have a recording? Audio files you upload (MP3, M4A, WAV, and more) are transcribed automatically and become searchable documents. Use Upload for those.

Starting from a prompt

When the editor is empty you will see a Start from a prompt button. Describe what you need and Lumen drafts it directly into the editor, ready for you to edit and publish. It is a starting point, not a separate place to work: the draft lands in the same editor and publishes the same way.

Quick TemplatesMeeting notes, how-to guides, templates, and project briefs
Custom PromptsDescribe anything and Lumen generates structured content
1

Open the prompt

With the editor empty, click Start from a prompt, then choose a quick template or type your own

2

Generate

Click Generate and Lumen writes the draft

3

Edit and publish

The draft lands in the editor behind the dialog. Edit as needed, then click Publish.

Toolbar tools

VoiceDictate text with your microphone instead of typing
Fix GrammarClean up spelling and grammar
ExpandAdd more detail to your content
SimplifyCondense long text into key points

Reviewing AI changes

After you click an Enhance option, a review window shows exactly what the AI changed: additions in green, removals struck through in red, plus a short summary of the changes. Click Accept to apply them or Reject to keep your original text. If the AI doesn't find anything worth changing, you'll see a message that the text is already clean instead of a silent change.

On some plans, a Custom Edit option is also available, letting you type your own instruction (for example, "make this more formal") instead of picking a preset.

Some knowledge does not sit still. A folder in Google Drive, a workspace in Notion, a board in Trello: the content changes, and a one-time upload goes stale the moment someone edits the original. Connect the tool instead and the documents stay current on their own.

1

Open Sync from an app

Click Sync from an app on the Add page to see the tools you can connect

2

Connect the tool

Sign in to the app and approve access, then choose which folders or spaces to bring in

3

Let it sync

Documents arrive on their own and refresh as the originals change. Watch progress from the Uploads tray on Add.

Pro Tip: Synced documents are managed by the source app. Their title and content are kept in step with the original, so edit them where they live rather than here. Your tags, access group, and summary stay yours to change.

Connecting a tool may depend on your plan and your role. See the Integrations article for the full list of apps and what each one brings across.

DocumentsPDF, DOC, DOCX, TXT, MD, RTF, HTML, Pages
SpreadsheetsXLS, XLSX, CSV, TSV, Numbers
PresentationsPPT, PPTX, Keynote
ImagesJPG, JPEG, PNG, GIF, WebP, BMP, TIFF, SVG
AudioMP3, M4A, WAV, OGG, FLAC, AAC, WMA, WebM
VideoMP4, MOV, AVI, WebM, MKV

Maximum file size: 50MB per file. No limit on number of documents.

After you add content, the system automatically processes it through several stages:

1

Text Extraction

Content is extracted from your file (OCR for images, transcription for audio/video).

2

Sectioning

Long documents are split into smaller, overlapping sections for better search accuracy.

3

Embedding

Each section is converted to a vector representation for semantic search.

4

Entity Extraction

AI identifies key people, organizations, concepts, and locations in your content.

5

Graph Integration

Entities and relationships are connected to your knowledge graph.

Important: Processed is not the same as answerable. If you upload without choosing an access group, the document finishes processing and appears in your Library under Pending Review, but Ask cannot use it and search will not return it. Open the Library, filter for Pending Review, and assign an access group to make it answerable. The same applies to files that arrive from a connected app such as Google Drive.

Processing typically takes 30 seconds to 2 minutes depending on document size. Monitor progress with the Uploads button on the Add page.

Use descriptive titlesGood titles help search and AI understand your content. "Q3 2024 Revenue Report" is better than "report.pdf."
Add tagsTags improve search accuracy and help organize your library. Use consistent tag names across documents.
Upload clean documentsRemove cover pages, table of contents, and boilerplate that add noise. The cleaner the content, the better the extraction.
Sync anything that keeps changingIf the original will be edited again, connect the app instead of uploading a copy. A copy goes stale; a synced document does not.
Check processing statusOpen Uploads on the Add page to ensure your documents completed successfully. Re-upload any that failed.
Upload fails immediatelyCheck file size (max 50MB), verify the file type is supported, and ensure you have a stable internet connection.
Processing stuckSome large files take longer to process. Open Uploads on Add to check status. If stuck for more than 10 minutes, try re-uploading.
URL fetch failsThe target website may block automated fetching. Try copying the text content and writing it up in Create instead.
The URL field disappearedStaging a file hides the URL field, because a single upload can only submit one thing. Remove the staged file to get the field back.
Poor search results after uploadVerify the document processed successfully. Try adding tags or a better title. For images, make sure you provided a good summary description.
Was this helpful?