Product docs

Add website, documents, videos, and cloud sources

AskAnyDocs can index website pages, sitemap URLs, exact URL lists, uploaded files, YouTube transcripts, Google Drive files, Notion pages, and manual Q&A pairs. This guide explains when to use each source and which file formats are supported.

Last updated: 2026-05-31

AskAnyDocs supports multiple knowledge source types so you can build a bot from the places where your product knowledge already lives. Choosing the right resource type keeps answers accurate and makes later troubleshooting easier.

Resource types at a glance

Resource type Best for
Website Discovering pages from a public website or one scoped path
Sitemap Importing structured URLs from an XML sitemap
URL list Indexing an exact list of known pages
File upload Adding documents, presentations, text files, HTML, or images
YouTube Indexing video transcripts from YouTube videos or playlists
Google Drive Connecting selected Drive files through OAuth
Notion Connecting selected Notion pages through OAuth
Q&A Adding exact question-and-answer pairs manually or from CSV

You can combine several resource types in one bot. A common setup is a website or sitemap for public docs, uploaded PDFs for policies, Google Drive or Notion for team knowledge, and Q&A pairs for short answers that should be exact.

Website resources

Use a website resource when you want AskAnyDocs to discover pages from a public site. If you enter the root domain, discovery scans the available sitemap and page structure. If you enter a scoped path such as /help or /docs, AskAnyDocs keeps the selection under that path.

Website resources are useful for:

  • public documentation sites
  • help centers
  • product pages
  • support article sections
  • onboarding guides

Review discovered pages before indexing. Exclude tag archives, login pages, old campaigns, duplicate pages, and content that does not help answer user questions.

Sitemap resources

Use a sitemap resource when you already know the XML sitemap URL. Sitemap import is usually the cleanest option for documentation platforms because it provides a structured list of canonical pages.

Sitemap resources work well when:

  • the site publishes a clean XML sitemap
  • you want to select pages before indexing
  • the documentation changes over time
  • you want predictable page discovery

If a sitemap contains too many unrelated pages, select only the sections that belong in the bot.

URL list resources

Use a URL list when you know the exact pages to index. Put each URL on its own line. This is the most controlled web-based source type and is a good choice for small launches, private beta docs, selected pricing pages, or a short list of high-value support articles.

Prefer final canonical URLs. Avoid redirects where possible because they make auditing harder.

File uploads

Use file uploads for content that is not available as web pages. AskAnyDocs currently accepts these upload extensions:

  • PDF: .pdf
  • Word: .docx
  • Rich text: .rtf
  • PowerPoint: .pptx
  • OpenDocument text: .odt
  • Plain text and Markdown: .txt, .md
  • CSV: .csv
  • HTML: .html, .htm
  • Images: .png, .jpg, .jpeg, .webp

Document files are split into searchable chunks and, where possible, into logical pages. Images are processed as image/OCR resources, so they work best when the text is clear and readable.

YouTube resources

Use a YouTube resource when important product knowledge exists in videos. AskAnyDocs resolves YouTube video or playlist URLs and indexes available transcript text for the bot's language when possible.

YouTube sources are useful for:

  • product walkthroughs
  • recorded onboarding videos
  • webinar clips that explain setup
  • training material with captions

Video indexing depends on transcript availability. If the video has no usable captions or transcript, create a text version of the content and upload it as a document or Q&A.

Google Drive resources

Google Drive sources let you connect selected Drive files through OAuth and sync them as bot knowledge. Drive is useful when your team maintains manuals, specs, policies, or onboarding files in Google Workspace.

Drive items can map to document-like source types such as PDF, DOCX, RTF, PPTX, ODT, HTML, or images depending on the file. Keep shared files organized and remove outdated versions before indexing.

Notion resources

Notion sources let you connect selected Notion pages and index their text content. This works well for internal knowledge bases, product runbooks, onboarding pages, and team process documentation.

Notion pages should be written with clear headings and direct answers. Large workspace dumps are usually less effective than a focused set of pages that match the bot's purpose.

Q&A resources

Q&A resources are for exact question-and-answer pairs. You can add pairs manually or import them from CSV.

Use Q&A when:

  • the answer should be short and exact
  • the content does not need a full article
  • support keeps answering the same question
  • you want a controlled response for a policy, limit, or setup detail

Q&A is not a replacement for full documentation, but it is effective for high-frequency questions.

Reindexing and syncing

Website, sitemap, URL list, Google Drive, and Notion sources can support reindexing or syncing depending on plan features. Uploaded files and Q&A are easier to control manually: replace or edit them when the source changes.

After updating sources, test the same questions users ask in the widget. If the bot still gives weak answers, improve the source text and reindex again.

Related articles