Q&A tables let you define explicit question-answer pairs. Each row becomes a retrievable chunk in the knowledge base — when a visitor asks something similar to one of your questions, the matching answer is retrieved and used by the AI.
Adding a Q&A Table
Open Knowledge Base tab
Go to your chatbot → Knowledge Base tab.
Click Add Source → Table
A table editor opens with Question and Answer columns.
Add entries
Type your questions and corresponding answers. Each row is independent.
Save
Click Save. Each Q&A pair is embedded as a separate chunk.
Editing Entries
Q&A tables can be edited in place. Open the table source, add/edit/delete rows, and save. Changes are re-embedded automatically.
When to Use Q&A Tables
| Use Case | Example |
|---|
| Known FAQs | ”What are your business hours?” → “We’re open Mon–Fri 9am–5pm.” |
| Precise answers | Questions where you want a specific, word-perfect response |
| Corrections | Override the AI’s general knowledge with your specific answer |
| Edge cases | Uncommon questions that other sources don’t cover |
Q&A Tables vs Other Sources
| Feature | Q&A Tables | Texts | Documents |
|---|
| Structured format | Yes (rows) | No (free text) | No (file) |
| Inline editing | Yes | Yes | No (re-upload) |
| Best for | Exact Q&A pairs | Longer content | Existing files |
| Chunk granularity | 1 chunk per row | Auto-chunked | Auto-chunked |
Q&A tables are best for precise, short answers. For longer content (explanations, guides, policies), use Texts or Documents instead — the auto-chunking produces better retrieval results.
