Integrate Semantic Search of Knowledge Base with Weaviate

[lalalune]

We want to add a lot of knowledge to our agents and have them be able to quickly search it. We are using Weaviate to do this already with Wikipedia, but we'd like to have our own trained knowledge base.

Here are the steps:

1. Set up Weaviate locally running in Docker container
2. Create API to create knowledge bases, add and remove documents to/from them
3. Create front end to create knowledge base and associate with each agent, with UI to add and remove documents
4. Use search to find relevant documents or information when speaker is asking the agent a question
5. Deploy weaviate to cloud or get cloud managed solution