ChatbotIQ supports multiple types of knowledge sources. You can mix and match — a single bot can draw from websites, PDFs, and manual Q&A entries at the same time.
| |
|---|
| What it is | A website URL that ChatbotIQ crawls to extract content. |
| How to add | Enter the base URL in the Source Wizard. ChatbotIQ discovers pages automatically. |
| What gets indexed | Text content from each page, optionally including images. Navigation, footers, and other UI elements can be stripped with Reader Mode. |
| Crawling modes | Standard (sitemap-based) or Advanced (link-following). See Configure Crawling. |
| Best for | Documentation sites, help centers, blogs, wikis, product pages. |
| Limitations | Pages behind login walls can’t be accessed. Very large sites may need plan upgrades for page limits. |
| |
|---|
| What it is | Uploaded PDF documents. |
| How to add | Click ”+ Add Source”, choose PDF, and upload one or more files. |
| What gets indexed | Extracted text from each page of the PDF. Each PDF page becomes one indexed page. |
| Best for | Product manuals, whitepapers, policy documents, reports. |
| Limitations | Scanned PDFs (images without OCR text) may have limited or no extractable content. |
| |
|---|
| What it is | Manually created question-and-answer pairs. |
| How to add | Click ”+ Add Source”, choose Q&A, then type in questions and answers. You can also bulk import from CSV or JSON. |
| What gets indexed | Each Q&A pair as a single unit. The question helps with retrieval; the answer is used in responses. |
| Best for | FAQs, specific answers that aren’t on your website, corrections to common misunderstandings, filling knowledge gaps. |
| Limitations | Manual maintenance required. No automatic updates. |
| |
|---|
| What it is | CSV or Excel files with structured data. |
| How to add | Click ”+ Add Source”, choose Spreadsheet, and upload your file. |
| What gets indexed | Rows are parsed and indexed as content. Column names provide context. |
| Best for | Product catalogs, pricing tables, feature comparison matrices, structured data. |
| Limitations | Very large spreadsheets may need to be split. Complex formatting is ignored. |
| |
|---|
| What it is | An OpenAPI 3.x YAML or JSON file describing your API. |
| How to add | Click ”+ Add Source”, choose OpenAPI, and upload your spec file. |
| What gets indexed | API endpoints, parameters, request/response schemas, and descriptions. |
| Best for | API documentation bots that help developers find and understand your endpoints. |
| Limitations | Only OpenAPI 3.x format is supported. |
Each bot can use multiple sources of different types. For example:
- A customer support bot might use your website (help center) + Q&A pairs (common questions) + PDFs (product manuals).
- An API docs bot might use your website (developer docs) + an OpenAPI spec (endpoint reference).
When a user asks a question, the bot searches across all assigned sources and uses the most relevant content regardless of source type.
