What you'll learn

Build efficient, practical applications including hybrid and multilingual searches, for diverse industries.
Understand vector databases and use them to develop GenAI applications without needing to train or fine-tune an LLM yourself.
Learn to discern when best to apply a vector database to your application.

Course recap

PRO

This course provides a practical introduction to vector databases using Weaviate, covering how to store, search, and manage vector embeddings for semantic search applications. It demonstrates automatic vectorization via OpenAI, CRUD operations, hybrid search combining sparse and dense vectors, and building RAG applications.

Concept map

Key concepts

Vector databasesSpecialized databases optimized for storing and querying high-dimensional vector embeddings
EmbeddingsDense vector representations generated by models like OpenAI's text-embedding; stored in the database for similarity search
Semantic searchFinding documents by meaning rather than keywords using vector similarity (cosine distance)
Hybrid searchCombining sparse vectors (keyword/BM25) with dense vectors (embeddings) for better retrieval
CRUD operationsCreate, Read, Update, Delete operations on vector database objects
Automatic vectorizationWeaviate's Text2Vec modules that generate embeddings during data ingestion
Filtered searchCombining vector similarity with metadata filters (e.g., category, date)

Lesson highlights

1.Introduction — Overview of vector databases and their role in AI applications
2.Vector Databases — Weaviate setup, automatic vectorization with OpenAI, semantic queries, CRUD operations
3.Sparse and Dense Vectors — Understanding sparse (keyword) vs. dense (embedding) representations
4.Hybrid Search — Combining BM25 and vector search with alpha parameter for result fusion
5.RAG with Vector Databases — Building question-answering systems with retrieved context
6.Generative Search — Using Weaviate's generative modules for LLM-powered responses
7.Multilingual Search — Cross-language retrieval with multilingual embeddings
8.Conclusion

About this course

Vector databases play a pivotal role across various fields, such as natural language processing, image recognition, recommender systems and semantic search, and have gained more importance with the growing adoption of LLMs.

These databases are exceptionally valuable as they provide LLMs with access to real-time proprietary data, enabling the development of Retrieval Augmented Generation (RAG) applications.

At their core, vector databases rely on the use of embeddings to capture the meaning of data and gauge the similarity between different pairs of vectors and sift through extensive datasets, identifying the most similar vectors.

This course will help you gain the knowledge to make informed decisions about when to apply vector databases to your applications. You’ll explore:

How to use vector databases and LLMs to gain deeper insights into your data.
Build labs that show how to form embeddings and use several search techniques to find similar embeddings.
Explore algorithms for fast searches through vast datasets and build applications ranging from RAG to multilingual search.

Who should join?

Anyone who’s interested in understanding and applying vector databases in their applications.

Vector Databases: from Embeddings to Applications

What you'll learn

Course recap

Concept map

Key concepts

Lesson highlights

About this course

Who should join?

Course Outline

Elevate your learning experience with Pro

Instructor

Sebastian Witalec

Vector Databases: from Embeddings to Applications

Want to learn more about Generative AI?