FAQ
Why is my AI Assistant not responding?
Check the following:
- Verify your OpenAI API key is correctly configured in Application Settings → AI Integration
- Ensure the API key has sufficient credits and is not expired
- Confirm the assistant status shows "Fully Synced" on the dashboard
- Check that the assistant is properly integrated in your Message Bot flows
How long does document processing take?
Document processing time depends on several factors:
- File Content: Documents with more text take longer to process
- File Size: Larger files require additional processing time
- Queue Load: Processing time may vary based on system queue
- Expected Time: 2-10 minutes per document
Wait for the "Sync Status" to show "Fully Synced" before testing your assistant.
What file formats are supported for the knowledge base?
The AI Assistant supports the following file formats:
- PDF Documents (.pdf) - up to 10MB
- Word Documents (.doc, .docx) - up to 10MB
- Text Files (.txt) - up to 10MB
- Markdown Files (.md) - up to 10MB
Ensure your files contain relevant, well-structured content for best results.
Why are my assistant responses inaccurate or irrelevant?
To improve response quality:
- Update Knowledge Base: Ensure uploaded documents contain current, relevant information
- Refine System Instructions: Adjust the assistant's behavior guidelines
- Check Document Quality: Use clear, well-structured content with proper headings
- Review Temperature Settings: Lower values (0.1-0.3) provide more focused responses
- Monitor Usage: Check the OpenAI Integration Details for performance insights
How do I integrate the AI Assistant with Message Bot flows?
Follow these steps:
- Navigate to Message Bot in the sidebar
- Click "Add Message Bot" to create a new bot
- Select "Personal Assistant" tab in Message Bot Options
- Choose your AI Assistant from the dropdown
- Configure trigger keywords and contact types
- Test the integration with sample queries
What should I do if document sync is stuck at 0%?
If documents aren't processing:
- Check file formats and ensure they're supported (.pdf, .doc, .docx, .txt, .md)
- Verify file sizes are under 10MB each
- Try re-uploading the documents
- Wait for the queue to process (it may take several minutes)
- Contact support if the issue persists after 30 minutes
How can I optimize costs for OpenAI API usage?
To manage API costs effectively:
- Choose Cost-Effective Models: Use GPT-4o Mini for basic queries, GPT-3.5 Turbo for general use
- Set Response Limits: Configure appropriate token limits (150-300 tokens)
- Monitor Usage: Regularly check the OpenAI Dashboard for usage statistics
- Optimize Content: Use clear, concise knowledge base documents
- Set Usage Alerts: Configure spending limits in your OpenAI account