A fun and absurd introduction to Vector Databases
summary
Hands-on demo building a search stack using vv8 for storage, a Spring Boot API and a Unity frontend that embeds text with a BERT-like model, images with ResNet50 and CLIP, and audio via Whisper, querying with HNSW with 3 neighbors and cosine distance (0 identical, 2 opposite). Demos index 50 weapons and ~721 Pokémon; text and spell icon lookups return sensible semantic clusters, image matching struggles with cartoon versus 3D views, and live metal-lyrics identification fails intermittently, illustrating fast reads, slow writes and lots of embedding trial-and-error.
AI-written summary · about
chapters (9)
- 0:00Vectors and their uses
- 7:02Vector databases explained
- 14:25Demo stack Weaviate Unity
- 21:08Semantic search weapons
- 25:39D&D thief spell search
- 30:46Pokémon image embeddings
- 36:35Spells linked to icons
- 41:57Shazam metal lyrics
- 46:22Q&A
related talks
48mA fun and absurd introduction to Vector Databases
48mA Fun & Absurd Introduction to Vector Databases
46mA Fun & Absurd Introduction to Vector Databases
46mA fun and absurd introduction to Vector Databases
50mA fun and absurd introduction to Vector Databases
41mGetting Started With Vector Databases
45mA gentle introduction to Vector Databases
01h 23mVector Databases Demystified: Discover how they work