Dev Tools · 2h ago
Developer Builds AI Knowledge Base with Aurora pgvector and Next.js
A developer created ChatScroll, an app that saves AI answers as searchable 'Scrolls' using Amazon Aurora PostgreSQL with pgvector for semantic search. The app stores 3072-dimensional vector embeddings and uses cosine similarity to find relevant results. It combines pgvector, ltree, and tsvector extensions for hybrid search across structured and chat data.
Meridian48 take
A practical demonstration of pgvector's capabilities, but the real test is whether such personal knowledge bases gain traction beyond hackathon projects.
Read the full reporting
How I Built a Personal AI Knowledge Base with Amazon Aurora pgvector and Next.js — AWS H0 Hackathon →
DEV Community
pgvectoramazon-aurora