Read the documentation.
Our SDK works hand in hand with our Notebook UI, where you can easily debug your chains after running relevance deploy
.
Start by installing our CLI:
npm install @relevanceai/chain -g
Then authenticate and fill in the project, region and API key:
relevance login
Then deploy your chains:
relevance deploy
We have a "step" for vector searching Redis! See the documentation.
See the example PDF Q&A chain in the chains
folder. It features in-flight vector search to intelligently populate LLM context.
Here's an extensive guide about using Redis + our Notebook UI to build a conversational chat experience.
Here's an example of building a SQL business analyst agent with our SDK.