Build Gen AI applications with Google Databases and Cloud Runtimes
summary
Prototype stores ~1,500 patent abstracts in AlloyDB, adds a 768‑dim vector column via AlloyDB with PGVector, calls Vertex AI embeddings (text-embedding-gecko-001/003) to convert text to vectors, and runs cosine‑similarity SQL queries to return nearest matches. A Java Cloud Function exposes a REST endpoint that accepts a query, converts it to an embedding, and returns top matches; end‑to‑end embedding + vector search completes in seconds and the demo deploys successfully from the provided GitHub repo.
AI-written summary · about
chapters (15)
- 0:00Introduction to Gen AI
- 3:51Databases and AI evolution
- 8:47Generative AI vs AI vs ML
- 11:17Vertex AI and databases
- 16:34Demo setup and overview
- 19:02Creating AlloyDB cluster
- 22:34Setting up database schema
- 24:17Enabling extensions
- 26:23Inserting patent data
- 27:05Testing embeddings
- 28:28Populating embeddings
- 29:12Granting Vertex AI access
- 30:17Vector search with SQL
- 33:30Deploying Cloud Function
- 36:02Testing the application
related talks
07mBuild Generative AI Apps for Enterprises with Cloud
02h 35mBringing GenAI to the Modern Enterprise. A production use-case. In Serverless Java!
09mGen AI for regulated industries
57mEnter The Brave New World of GenAI with Vector Search
08mGenerative AI app development with Cloud SQL for PostgreSQL
01h 16mAzure SQL ❤️ Open AI
14mBring AI to Your Data
27mUse AI Vector Search to Build GenAI Apps with Enterprise Data