Understanding Vector Databases Online Course

Understanding Vector Databases Online Course

This course provides a comprehensive introduction to vector databases, starting with core concepts and their growing role in modern data management. You’ll learn the differences between traditional and vector databases, explore leading solutions like Chroma and Pinecone, and gain hands-on experience setting up environments, creating and querying databases, and applying similarity measures such as cosine similarity and Euclidean distance. The course also covers integration with Large Language Models (LLMs) and LangChain, guiding you through workflows from document loading to response generation. By the end, you’ll have both a strong conceptual foundation and practical skills to choose, implement, and leverage vector databases for advanced data management tasks.

Who should take this course?

This course is ideal for developers, data scientists, AI/ML practitioners, and database professionals who want to understand how vector databases work and how they power modern AI applications. It’s well-suited for those with a basic understanding of data structures or machine learning who are interested in semantic search, recommendation systems, embeddings, and large-scale AI-driven solutions. Whether you’re building intelligent applications, working with unstructured data, or exploring the backbone of generative AI systems, this course will equip you with the knowledge to effectively use vector databases.

What you will learn

• Grasp the fundamental concepts and applications of vector databases.
• Analyze the differences between traditional and vector databases.
• Set up and use vector databases like Chroma and Pinecone.
• Apply vector similarity measures for advanced data analysis.
• Use Large Language Models and frameworks like LangChain with vector databases.
• Understand the fundamentals of vector databases and their importance in data management.

Course Outline

Introduction

• Introduction - Course prerequisites and structure

Vector Databases Deep Dive - Fundamentals

• Introduction to Vector Databases - Full Overview
• Why Vector Databases
• Vector Databases - Benefits and Advantages

Traditional vs Vector Databases - Differences

• Traditional vs Vector Databases - Limitations and Challenges
• Vector Databases & Embeddings - Full Work Flow
• Embeddings vs Vectors – Differences
• Vector Databases - How They Work and Advantages
• Vector Databases Use Cases
• Vector and Traditional Databases - Summary

Vector Databases Solutions - Top 5 Vector Databases

• The Top 5 Vector Databases - Overview
• Understanding LLM (Large Language Models)

Building Vector Databases - Hands-on - Chroma Vector Database

• Development Environment Setup
• Setup VS-Code, Python and OpenAI API Key
• Chroma Database workflow
• Creating a Chroma Vector Database & Adding Documents & Querying them
• Looping Through the Results & Showing Similarity Search Results
• Chroma Default Embedding Function
• Chroma Vector Database - Persisting Data and Saving
• Creating an OpenAI Embeddings - Raw without Chroma
• Using OpenAI's Embedding API to Create Embedding in Chroma
• Vector Databases Metrics and Data Structures
• Section Summary

Common Measures of Vector Similarity

• Vector Similarity Deep Dive - Cosine Similarity
• Euclidean Distance - L2 Norm
• Dot Product
• Section Summary

Vector Databases and LLM - the Full Workflow

• Vector Databases and LLM - Deep Dive
• Loading all Documents
• Generating Embeddings from Documents & Insert them into Chroma Database
• Getting the Relevant Chunks when Given a Query
• Using OpenAI LLM to Generate Response - Full Flow
• Section Summary

Vector Databases & the Langchain Framework

• The LangChain Framework - Quick Overview
• Getting started with LangChain and the OpenAIChat Wrapper
• Loading Documents with LangChain Document Loader
• Splitting the Documents with LangChain
• Creating a Chroma Vector Database with LangChain
• Getting the Response from the Model - the Complete Workflow

Pinecone Vector Database

• Pinecone - Deep Dive
• Create Pinecone Account & Dashboard Overview
• Creating our Pinecone Index in Code
• Upserting and Querying our Pinecone Index
• Querying Pinecone Manually in the Dashboard
• Using LangChain Pinecone Wrapper - Create Index and Upsert & Similarity Search
• Creating a Retriever and Chain Objects & a LLM to get a Response
• Clean up - Delete Pinecone Index
• Challenge - Explore other Vector Database
• Section Summary

Choosing the Right Vector Database

• Choosing the Right Vector Database - Comparison Tables
• Which Database Should I Choose?
• Choosing the Right Database - Criteria

Wrap up & Next Steps

• Congratulations and Next Steps