NeurondB
A complete AI solution
PostgreSQL extensions for production use
Vector search uses HNSW and IVF indexing.
ML inference runs ONNX models. Generates text, image, and multimodal embeddings.
Hybrid search combines semantic and full-text search. Cross-encoder reranking. RAG pipeline.
GPU acceleration supports CUDA and ROCm. Faster matrix operations.
Products
PostgreSQL extensions and tools for AI, high availability, connection pooling, monitoring, and multi-protocol support.
NeuronDB
AI database extension with vector search, RAG pipeline, ML inference, and GPU acceleration.
- •Vector search with HNSW indexing
- •52 ML algorithms
- •473 SQL functions
- •GPU acceleration (CUDA/ROCm/Metal)
- •RAG pipeline in-database
pgraft
Raft consensus for PostgreSQL with automatic leader election and split-brain prevention.
- •Automatic leader election
- •Split-brain prevention
- •Zero-downtime failover
- •etcd-io/raft based
- •Background worker architecture
pgBalancer
Connection pooler with load balancing, REST API, and machine learning query routing.
- •AI-powered load balancing
- •REST API (17 endpoints)
- •MQTT event streaming
- •Connection pooling
- •YAML configuration
FauxDB
Dual-protocol database supporting MongoDB and MySQL wire protocols on PostgreSQL.
- •MongoDB wire protocol
- •MySQL wire protocol
- •PostgreSQL ACID guarantees
- •SQL query translator
- •Rust-powered performance
pg_stat_insights
Performance analytics extension with query analysis and optimization recommendations.
- •Query performance analysis
- •Table and index statistics
- •Cache hit ratio monitoring
- •Optimization recommendations
- •52 metrics across 11 views
pgSentinel
Monitoring platform with Grafana dashboards, Prometheus integration, and real-time metrics.
- •Grafana dashboards
- •Prometheus integration
- •Real-time metrics
- •Alerting system
- •Docker deployment