conscious-sapphire•2y ago
Latency with Replit integration
First time builder, and I used Replit for my knowledge base and the AI integration. However, my chatbot is extremely slow in answering (i have the latency set at 100). Could this be due to Replit? Due to the fact that I am on a free plan? Any other setting I should be looking at?
2 Replies
The latency you're experiencing with your chatbot could be due to several factors. While it's possible that the free plan or Replit could be contributing to the delay, it's also important to consider other factors such as the complexity of your knowledge base and the AI model you're using.
In Voiceflow, different AI models have different costs and performance characteristics. For instance, more accurate models like GPT-4 use more tokens and could potentially result in slower response times. Also, the size and complexity of your knowledge base could affect the speed of responses. A large knowledge base with complex queries could take longer to process.
Another factor to consider is the 'token' slider on the AI steps. This controls the output message and depending on your design, there may be a large amount of information that is sent to the AI model as part of your prompt.
Lastly, the 'No Reply Response' feature in Voiceflow allows you to set a timeout period for user input. If this is set to a high value, it could also contribute to perceived latency.
For more detailed information, you can refer to the following documents:
- Voiceflow AI Token Documentation
- Voiceflow No Reply Response Documentation
- Voiceflow Live Chat Handoff Documentation
I presume your connecting with the assistant API, and if thats the case it’s not replit it’s just very slow! 🐌