Using Postgres + OpenAI to power your AI Recommendation...
summary
Claims Postgres alone can run a recommendation engine by storing and querying OpenAI embeddings and avoiding a separate vector DB. Uses Rails + ActiveRecord to call OpenAI embeddings and persist vectors into Postgres + pgvector, performs cosine/ANN search inside the database, and returns low-latency (subsecond) recommendations. Shows a working prototype deployed with batching, client-side and DB caching, index tuning, partitioning and read-replica strategies for scale, and reports cost and latency advantages versus adding an external vector service.
AI-written summary · about
chapters (5)
- 0:00Introduction
- 13:08Setting up Postgres
- 24:46Integrating OpenAI
- 37:27Building recommendations
- 44:20Q&A and closing
related talks
44mPostgres and OpenAI - a match made in Azure
27mPutting AI in Production
47mExtendable by Design: Building Generative AI Apps with Postgres and Vector Storage and Azure AI
44mIntegrating AI with Postgres...
34mAnalytics and AI with PostgreSQL - Mission Possible!
50mFrom Transactions to Intelligence: Postgres for Open AI & Analyti
53mModern PostgreSQL for AI...
12mThe mechanics of integrating AI and vector search into Azure Database for PostgreSQL