Rivestack

Rivestack, built by a France-based team with infrastructure across EU Central, US-East, and Singapore, provides managed PostgreSQL databases optimized for AI workloads and specifically tuned for vector search using the pgvector extension. It enables semantic search and retrieval-augmented generation (RAG) capabilities without requiring a separate, dedicated vector database. Rivestack offers dedicated PostgreSQL on NVMe storage with pre-tuned pgvector and HNSW indexing, sub-4ms latency for vector queries, a table editor for browsing and managing data without SQL knowledge, and an AI-assisted SQL editor for query generation. It includes Terraform integration for infrastructure-as-code provisioning, built-in embeddings generation via OpenAI's text-embedding-3-small model, and high-availability support with Patroni failover and point-in-time recovery. Pricing uses fixed monthly rates with no per-query or egress charges: a Free tier offers shared 256MB RAM, Solo is $29/month (1 vCPU, 2GB RAM), Starter is $49/month (2 vCPU, 4GB RAM), Growth is $85/month (4 vCPU, 8GB RAM), and Scale is $159/month (8 vCPU, 16GB RAM).