Weaviate Academy
by Weaviate
A sequenced 17-hour path from 'what is a vector index' to a working search-and-RAG API.
Overview
Weaviate Academy is the vendor's own structured curriculum for its open-source vector database, now served as a standalone site at academy.weaviate.io (the old docs path redirects there). Courses carry codes and stated runtimes: WA050 'Key Concepts & Architecture' (Beginner, 1 hour, three modules - What is Weaviate?, Core concepts, System architecture and use cases), WA101T-PY 'Hands-on Weaviate with Python' (Beginner, 1 hour, covering data storage, search and RAG), WA150-PY 'A Gentle Deep Dive into AI Models' (Intermediate, 2 hours), WA180-PY 'Your First AI App (Search and RAG)' (Intermediate, 3 hours, building a movie-recommendation API on Weaviate plus FastAPI across two modules and five endpoints, with WA050 and WA101T listed as prerequisites) and WA210-PY 'Architectural Decisions Guide' (1.5 hours). These are sequenced into a 17-hour 'AI Developer' learning path split into Prototyping (7 hours), Development (7 hours - Embedding Model Evaluation & Selection, Search Strategies: In Depth, Data Ingestion: In Depth) and Additional Standalone Topics (3 hours - Multimodal Data: In Focus, Document Chunking Strategies). An 'Introduction to Weaviate Agents' course is listed as Coming Soon. Lesson pages render in full without payment or signup; login only tracks progress, and the site states that courses are for educational purposes with certification issuance still in development. The material is written by the engineering and developer-relations team maintaining Weaviate itself - a BSD-3 licensed Go database with roughly 16.7k GitHub stars and daily commits - so the client APIs shown track the shipped product rather than a frozen snapshot.
At a Glance
- Topic
- RAG
- Level
- All Levels
- Format
- Course
- Cost
- Free
- Duration
- ~17 hours total, self-paced (individual courses 1-3 hours each)
- Provider
- Weaviate
- Hands-on
- Yes — code/exercises
- Certificate
- None
What You’ll Learn
- ✓Model Weaviate collections, objects, vectors and indexes as one coherent data model
- ✓Store text, your own precomputed vectors, and multimodal data via the Python client
- ✓Build a movie-recommendation REST API with FastAPI backed by Weaviate search
- ✓Run retrieval-augmented generation against a collection using generative model integrations
- ✓Evaluate and select embedding models instead of defaulting to whichever comes first
- ✓Compare keyword, vector and hybrid search strategies and judge when each one wins
- ✓Apply vector compression and multi-tenancy to keep memory cost sane at scale
- ✓Choose document chunking strategies and reason about how they change retrieval quality
Highlights
- •Sequenced into one 17-hour 'AI Developer' path rather than a pile of unordered tutorials
- •Every course publishes an explicit runtime, level and prerequisite course codes up front
- •Genuinely free with no paywall - login exists only to track progress, no card required
- •Written by the team maintaining the ~16.7k-star Weaviate database, so client code stays current
- •Goes past the demo into architectural decisions, vector compression and multi-tenancy
- •Capstone ships an actual FastAPI service with five endpoints, not a notebook cell
Who It’s For
Best For
- ✓Backend engineers adding semantic search to an existing product
- ✓AI engineers building their first production RAG pipeline
- ✓Teams evaluating Weaviate against other vector databases before committing
- ✓Developers who want a sequenced path instead of assorted blog tutorials
Prerequisites
- •Working Python - the hands-on courses are Python-based (course codes end in -PY)
- •Comfort with REST APIs and JSON; the capstone course is built on FastAPI
- •Access to a model provider such as OpenAI for the embedding and generative modules
- •No prior vector-database experience needed - WA050 starts from first principles
FAQ
What is Weaviate Academy?
Weaviate Academy is the free, self-paced course track run by the team behind the Weaviate vector database. It takes you from what a vector index is through a working search-and-RAG application, and on to the decisions that bite in production - embedding model choice, chunking, compression and multi-tenancy. Finish it and you can stand up a Weaviate collection, load your own data, run hybrid search and serve RAG results from a FastAPI service.
Is Weaviate Academy free?
Weaviate Academy is free to access.
What level is Weaviate Academy for?
Weaviate Academy is aimed at a all levels audience. Recommended background: Working Python - the hands-on courses are Python-based (course codes end in -PY), Comfort with REST APIs and JSON; the capstone course is built on FastAPI, Access to a model provider such as OpenAI for the embedding and generative modules, No prior vector-database experience needed - WA050 starts from first principles.
How long does Weaviate Academy take?
Expect roughly ~17 hours total, self-paced (individual courses 1-3 hours each). Most learners work through it at their own pace.
What will I learn from Weaviate Academy?
You'll learn: Model Weaviate collections, objects, vectors and indexes as one coherent data model; Store text, your own precomputed vectors, and multimodal data via the Python client; Build a movie-recommendation REST API with FastAPI backed by Weaviate search; Run retrieval-augmented generation against a collection using generative model integrations; Evaluate and select embedding models instead of defaulting to whichever comes first; Compare keyword, vector and hybrid search strategies and judge when each one wins; Apply vector compression and multi-tenancy to keep memory cost sane at scale; Choose document chunking strategies and reason about how they change retrieval quality.
Topics
Sources
This page was written from 6 sources, 2 on domains other than academy.weaviate.io.