Vector Databases in Practice: Deep Dive

This is the repository for the LinkedIn Learning course Vector Databases in Practice: Deep Dive. The full course is available from LinkedIn Learning.

Vector databases and their uses are transforming how we work. They are fundamentally changing how data is stored, managed, and retrieved through their deep integration with AI models. In this course, learn practical, end-to-end skills on how to build and use vector databases. Instructor Joon-Pil Hwang guides you through building an application that is primarily powered by a vector database, taking you all the way from database creation to usage and even app integration. Learn key considerations in using a vector database in practice, as well as be aware of some common techniques and baseline choices as starting points. Discover keyword, vector, and hybrid searches to find the right data faster, as well as how to apply retrieval-augmented generation - which makes generative AI tools more accurate by grounding them with your data.

See the readme file in the main branch for updated instructions and information.

Instructions

You will be using Python scripts with .py extensions for this course. Each script is named to correspond to a video - such as 02_04_import.py for video 02_04.

Installation

1. To use these exercise files, you must have the following installed:
   • Python 3 (3.10 or newer recommended)
   • Python libraries:
     • Weaviate Python client (Latest available version recommended, developed with 4.4.3)
     • Pandas (Developed with 2.1.4)
     • Streamlit (Developed with 1.29.0)
     • MediaWikiAPI (Developed with 1.2.0)
2. Clone this repository into your local machine using the terminal (Mac), CMD (Windows), or a GUI tool like SourceTree.
3. You will need your own Weaviate database instance. Section 2.01 will show you how to do so.
   • As an alternative, you can run Weaviate any way you prefer, such as through Docker or Kubernetes. Please see the official documentation for more details.

Useful links

Documentation & resources

• Weaviate documentation
• Weaviate Python client overview
• Weaviate Python client documentation
• Weaviate Cloud Services
• Streamlit documentation

Blogs, articles & resources on AI

• The Weaviate blog
  • A Gentle Introduction to Vector Databases
• Hugging Face blog
• Cohere blog
• Towards Data Science
• Google AI blog
• ChatGPT Is a Blurry JPEG of the Web

Instructor

Joon-Pil Hwang

Technical Curriculum Developer at Weaviate

Check out my other courses on LinkedIn Learning.