FromZero2AI
← Back to Courses
Intermediate~10 hours8 lessons

Embeddings & Vector Databases

Understand how text becomes numbers, how vector databases store and search meaning, and build a semantic search pipeline from scratch.

By Bishwambhar Sen • Last updated

What you'll build: a high-performance vector similarity search engine
Start Course

Curriculum

Module A: Understanding Embeddings
3 lessons

Lesson 1: What Are Embeddings? Text as Vectors

Understand how embedding models map words, sentences, and documents to points in high-dimensional space — and why proximity in that space means semantic similarity.

60 min

Lesson 2: Generating Embeddings with APIs

Call the OpenAI and Anthropic embedding APIs, inspect the raw vectors, and compute cosine similarity between sentences by hand.

75 min

Lesson 3: Choosing the Right Embedding Model

Compare embedding model families by dimension count, context length, cost, and benchmark performance. Know when to use a hosted API vs. a local model like nomic-embed.

55 min
Recommended Watch

Attention in transformers, visually explained

Deep dive into the attention mechanism and how transformers create meaningful embeddings from text.

3Blue1Brown26:20
Module B: Vector Databases
3 lessons

Lesson 4: Vector Database Fundamentals

Learn what a vector database is, how it differs from a relational database, and get hands-on with ChromaDB — the most learner-friendly vector store.

70 min

Lesson 5: Similarity Search: ANN & Distance Metrics

Explore cosine similarity, dot product, and Euclidean distance. Understand why Approximate Nearest Neighbor (ANN) search is essential at scale.

65 min

Lesson 6: HNSW Indexing & Performance Trade-offs

Understand the Hierarchical Navigable Small World (HNSW) algorithm that makes billion-scale vector search feasible. Learn the recall vs. speed trade-off.

55 min
Recommended Watch

Vector Databases simply explained!

What vector databases are, how they store embeddings, and why they're essential for AI search.

IBM Technology8:26
Module C: Building Semantic Systems
2 lessons

Lesson 7: Building a Semantic Search Pipeline

Wire together an embedding model and a vector database into a complete semantic search system: ingest documents, embed them, and retrieve by meaning — not keywords.

80 min

Lesson 8: Capstone — Document Intelligence System

Build an end-to-end document intelligence system that ingests a PDF, chunks it, embeds every chunk, stores them in ChromaDB, and answers natural language questions over the content.

120 min
Recommended Watch

Semantic Search with Vector Databases

Build a semantic search system from scratch using embeddings and vector similarity.

James Briggs14:35

Course Quiz

5 questions · ~5 min · Pass 4/5 to unlock your badge

Start Quiz →