Tutorial: AI/ML, Vector Search & RAG

Level: Advanced | Time: 15 minutes | Version: 7.2.0

HeliosDB includes built-in vector search, RAG pipelines, NL2SQL, and embedding support.

Vector Search

-- Check vector index status
SHOW VECTOR INDEXES;
-- Engine: HNSW + IVF, SIMD-accelerated

-- Check available embeddings
SHOW EMBEDDINGS;
-- Providers: OpenAI, Cohere, Voyage, Local

-- Full-text search indexes
SHOW FULLTEXT INDEXES;

RAG Pipelines

-- View RAG pipeline components
SHOW RAG PIPELINES;
-- Chunking: fixed, paragraph, sentence, sliding, semantic
-- Retrieval: vector, fulltext, hybrid, reranker
-- LLM: openai, anthropic, local
-- Graph RAG: knowledge graph + retrieval

Natural Language to SQL

-- NL2SQL engine status
SHOW NL2SQL STATUS;
-- Engine: heliosdb-nl
-- Providers: openai, anthropic, deepseek, grok
-- Features: multi-turn, schema-aware, few-shot, agentic

Cluster & Federation

-- Raft consensus status
SHOW RAFT STATUS;

-- Cluster nodes
SHOW CLUSTER NODES;
SHOW CLUSTER HEALTH;

-- Federation (cross-cloud FDW)
SHOW FEDERATION;
-- S3, Azure Blob, Google Cloud Storage wrappers

-- Lakehouse catalogs
SHOW CATALOGS;
-- Hive Metastore, Nessie, AWS Glue, Azure, Unity

Advanced Operations

-- Autonomous index management
SHOW INDEX RECOMMENDATIONS;
SHOW AUTONOMOUS INDEXES;
APPLY INDEX RECOMMENDATION idx_001;

-- Backup & recovery
SHOW BACKUP STATUS;
SHOW REPLICATION STATUS;

-- Data integrity
SHOW INTEGRITY STATUS;

-- Snapshot management
CREATE SNAPSHOT before_migration;
SHOW SNAPSHOTS;
DROP SNAPSHOT before_migration;

-- WASM functions
SHOW WASM FUNCTIONS;
CREATE WASM FUNCTION my_transform AS '/path/to/transform.wasm';
DROP WASM FUNCTION my_transform;

-- Streaming / CEP
SHOW STREAMS;

-- Scaling
SHOW SCALING STATUS;
SHOW ENERGY STATUS;

-- Data lineage (blockchain-backed)
SHOW LINEAGE orders;

-- Materialized views
SHOW MATERIALIZED VIEWS;
REFRESH MATERIALIZED VIEW monthly_stats;