So we are using WatsonX (with prompt) with Assistant (using a RAG model to Discovery); what is the advantage here using flow and or what is the difference. We are exploring the most cost-effective way to do RAG with deployed white-labeled assistants.
@@ramonaortega1975flows engine is aimed at developers and sits somewhere between “code everything from scratch” and low-code/no-code experience in the browser. If allows for adding customizations and a lot of abstractions for common AI patterns, only when you want to add a lot of customizations you need to write some (declarative) code. One big plus it spins up its own infrastructure.
Yes you can! Use the endpoint printed in your terminal after using the deploy function, and set a header for “authorization” with the value “apikey YOUR_APIKEY”.
Hope you enjoyed this video! Please let us know what topics about AI you’d like to see videos on!
So we are using WatsonX (with prompt) with Assistant (using a RAG model to Discovery); what is the advantage here using flow and or what is the difference. We are exploring the most cost-effective way to do RAG with deployed white-labeled assistants.
@@ramonaortega1975flows engine is aimed at developers and sits somewhere between “code everything from scratch” and low-code/no-code experience in the browser. If allows for adding customizations and a lot of abstractions for common AI patterns, only when you want to add a lot of customizations you need to write some (declarative) code. One big plus it spins up its own infrastructure.
Great content, well explained!
Thank you 🎉
Thank you for such an interesting walkthrough. Can you please help me to inference this endpoint using Postman?
Yes you can! Use the endpoint printed in your terminal after using the deploy function, and set a header for “authorization” with the value “apikey YOUR_APIKEY”.
It’s GraphQL what you can use to directly use the endpoint, Postman can work with this