Quick Start
Response Quality
Improve how accurately and helpfully your chatbot answers customers.
Documentation Index
Fetch the complete documentation index at: /llms.txt
Use this file to discover all available QwryAI pages before exploring further.
Response quality is the result of source coverage, retrieval quality, model behavior, and clear escalation rules. Review weak answers with the customer question, the source content, and the chatbot response side by side.
- Improve missing content first.
- Use Playground to retest the same wording after every change.
Quality Levers
Instructions
Clear instructions tell the chatbot how to answer, when to ask for more information, and when to escalate. Keep instructions specific to your business and avoid broad promises the source content cannot support.
Source quality
The best answer improvements usually come from clearer source content. Add missing policy details, rewrite vague FAQs, and remove outdated pages before changing model settings.
Review loop
- Find weak answers in conversations, analytics, or knowledge gaps.
- Compare the customer question with the available source content.
- Update the source or instruction that caused the weak answer.
- Retest the exact wording in Playground.
- Monitor future conversations for the same issue.
Create a URL Mapping Document
When customers ask for a page by purpose instead of name, a short URL mapping document helps the chatbot recommend the right destination. Keep the file simple: page name, URL, and when to suggest it.
| Page | URL | When to suggest |
|---|---|---|
| Returns policy | https://example.com/returns | Customer asks about return windows, refunds, or exchanges. |
| Shipping information | https://example.com/shipping | Customer asks about delivery regions, timelines, or fees. |
| Contact support | https://example.com/contact | Customer needs a human team or account-specific help. |