Vector Databases in Practice: Deep Dive

Go beyond the basics of vector databases by building a database and app from scratch, and learn key considerations along the way.

Introduction

• The power of AI-powered vector databases

1. Search Functions in a Vector Database

• A high-level view of vector databases
• What you can do with vector databases
• Get set up for the course
• Keyword filtering and keyword searches
• Vector searches
• Searching with filters
• Hybrid searches
• Retrieval augmented generation
• Challenge: Vector database queries
• Solution: Vector database queries

2. Building a Vector Database

• Create your own database
• Work with Weaviate
• Create an object collection
• Basic data import in Weaviate
• Establishing relationships with references
• Recap: Building a vector database
• Challenge: Add another object collection
• Solution: Add another object collection

3. Building a Vector Database-Powered App

• Web apps and vector databases
• Create a basic app
• Connect the app to Weaviate
• Parsing query responses
• Recommendations with RAG
• Challenge: App enhancements
• Solution: App enhancements

4. Making a Vector Database Work for Your Data

• Messiness of real data
• Pre-processing text for vector databases
• Chunking longer texts
• Chunk Wikipedia articles
• Challenge: Import Wikipedia data chunks
• Solution: Import Wikipedia data chunks

Conclusion

• Continue learning about vector databases