What are knowledge sources?
Knowledge sources are the content your bot learns from. When a visitor asks a question, the bot searches your knowledge base for relevant information and uses it to generate an accurate answer.
Your bot only answers based on what's in its knowledge base — it won't make up information.
Source types
URLs
Enter any public URL and AnsrNow will crawl the page to extract its text content. You can also enable multi-page crawling to automatically follow links and index subpages from the same domain.
- The Free plan supports up to 200 pages per source
- The Pro plan supports up to 500 pages per source
- Only pages on the same domain are crawled
- JavaScript-rendered content (SPAs) may not be fully captured
PDF files
Upload PDF documents directly. AnsrNow extracts the text, splits it into chunks, and indexes it for search. This is ideal for:
- Product manuals and handbooks
- Policy documents
- Training materials
- Research papers or reports
Tip: For best results, use text-based PDFs rather than scanned images. If your PDF is a scanned document, the text extraction may be limited.
How content is processed
When you add a knowledge source, AnsrNow processes it in three steps:
- Extraction — Text is extracted from the URL or document.
- Chunking — The text is split into smaller, overlapping chunks (typically a few hundred words each) so the bot can find specific, relevant passages.
- Embedding — Each chunk is converted into a numerical vector (embedding) that captures its meaning. These embeddings are stored and used for similarity search when a visitor asks a question.
Processing happens in the background and usually completes within a few minutes, depending on the size of your content.
Checking source status
Each knowledge source shows its processing status on the Knowledge Sources page:
- Processing — Content is being crawled and indexed
- Completed — Ready to use, all documents have been indexed
- Failed — Something went wrong (e.g., the URL is unreachable or the file is corrupted)
You can also see the number of documents (chunks) created from each source.
Reprocessing sources
If your content has changed — for example, you've updated your help docs or published new pages — you can reprocess a knowledge source to refresh the bot's knowledge. Click the Reprocess button next to any source to re-crawl and re-index it.
Limits by plan
- Free — 1 knowledge source per bot, up to 200 pages per URL source
- Pro — 20 knowledge sources per bot, up to 500 pages per URL source
Next: Chat widget & standalone page →