Introduction to AI-Native Vector Databases
2h 48mIntermediate2023-12-14
Authors
Zain Hasan
Course details
The primary purpose of vector databases is to provide fast and accurate similarity search or nearest neighbor search capabilities. The integration of AI techniques in vector databases enhances their capabilities, improves search accuracy, optimizes performance, and enables more intelligent and efficient management of high-dimensional data. In this course, Zain Hasan introduces this foundational technology—which is already being used in industries like ecommerce, social media, and more. Zain covers everything from foundational concepts around AI-first vector databases to hands-on coding labs for question answering using LLMs.
Skills covered
Natural Language Processing (NLP)Introduction toMachine LearningDatabase DevelopmentArtificial Intelligence FoundationsDatabase ManagementArtificial Intelligence (AI)Software Development
Concepts
0. Introduction
- 01 - Learning AI-native vector databases
- 02 - What you should know
- 03 - The superpower of vector databases
1. Data - What Data Do Vector Databases Store and How Is It Stored
- 04 - Structured versus unstructured data
- 05 - Human-understandable versus machine-understandable data
- 06 - Drawing out and visualizing vector representations of data
- 07 - Introduce the concept of distance between two vectors
- 08 - Challenge - Working with vectors
- 09 - Solution - Working with vectors
2. Natural Querying - How Do You Search for Data in a Vector Database
- 10 - Frame the query as a question or search
- 11 - Generate the question in machine-understandable language
- 12 - Adding data to a vector database
- 13 - Performing semantic searches using Weaviate
- 14 - Challenge - Vector search with Weaviate
- 15 - Solution - Vector Search with Weaviate
3. Machine Learning Vectors - How Does a Vector Database Understand Your Data
- 16 - Machine learning models and object classification
- 17 - Translating data from human to machine-understandable
- 18 - ML models and vector embeddings
- 19 - Challenge - Search with images and text
- 20 - Solution - Search with images and text
4. Scalability - What Does a Vector Database Need to Do
- 21 - Scalability - When to use a vector DB
- 22 - Ways to measure performance of a vector DB
- 23 - CRUD operations in vector DBs
- 24 - Challenge - CRUD and performance
- 25 - Solution - CRUD and performance
5. Demonstrate Vector DBs and Use Cases
- 26 - Vector DB1 - E-commerce RecSys
- 27 - Vector DB2 - Hybrid search
- 28 - Vector DB3 - Retrieval augmented generation
- 29 - Challenge - Vector DBs
- 30 - Solution - Vector DBs
Conclusion
- 31 - Continue your AI-native vector databases learning journey