CREATE EXTENSION IF NOT EXISTS vector;

CREATE TABLE documents (
    id SERIAL PRIMARY KEY,
    content TEXT NOT NULL,
    embedding vector(1536)  -- text-embedding-3-small
    -- Or: embedding halfvec(3072)  -- text-embedding-3-large (50% memory)
);

SELECT id, content, 1 - (embedding <=> query_vec) AS similarity
FROM documents
ORDER BY embedding <=> query_vec
LIMIT 10;

CREATE INDEX ON documents USING hnsw (embedding vector_cosine_ops);

# pgvector with PostgreSQL 17
docker run -d --name pgvector-db \
  -e POSTGRES_PASSWORD=postgres \
  -p 5432:5432 \
  pgvector/pgvector:pg17

# Or PostgreSQL 18 (latest)
docker run -d --name pgvector-db \
  -e POSTGRES_PASSWORD=postgres \
  -p 5432:5432 \
  pgvector/pgvector:pg18

# ParadeDB (includes pgvector + pg_search + BM25)
docker run -d --name paradedb \
  -e POSTGRES_PASSWORD=postgres \
  -p 5432:5432 \
  paradedb/paradedb:latest

embedding <=> query  -- Cosine distance (1 - similarity)
embedding <-> query  -- L2/Euclidean distance
embedding <#> query  -- Negative inner product

-- Top 10 similar (cosine)
SELECT * FROM docs ORDER BY embedding <=> $1 LIMIT 10;

-- With similarity score
SELECT *, 1 - (embedding <=> $1) AS similarity FROM docs ORDER BY 2 DESC LIMIT 10;

-- With threshold
SELECT * FROM docs WHERE embedding <=> $1 < 0.3 ORDER BY 1 LIMIT 10;

-- Preload index (run on startup)
SELECT 1 FROM docs ORDER BY embedding <=> $1 LIMIT 1;

-- HNSW (recommended)
CREATE INDEX ON docs USING hnsw (embedding vector_cosine_ops);

-- With tuning
CREATE INDEX ON docs USING hnsw (embedding vector_cosine_ops)
WITH (m = 24, ef_construction = 200);

-- Query-time recall
SET hnsw.ef_search = 100;

-- Iterative scan for filtered queries (pgvector 0.8+)
SET hnsw.iterative_scan = relaxed_order;
SET ivfflat.iterative_scan = on;

Query type?
├─ Conceptual/meaning-based → Pure vector search
├─ Exact terms/names → Pure keyword search (FTS)
├─ Fuzzy/typo-tolerant → pg_trgm trigram similarity
├─ Autocomplete/prefix → pg_trgm + prefix index
├─ Substring (LIKE/ILIKE) → pg_trgm GIN index
└─ Mixed/unknown → Hybrid search
    ├─ Simple setup → FTS + RRF (no extra extensions)
    ├─ Better ranking → BM25 + RRF (pg_search extension)
    └─ Full-featured → ParadeDB (Elasticsearch alternative)

Document count?
├─ < 10,000 → No index needed
├─ 10k - 1M → HNSW (best recall)
└─ > 1M → IVFFlat (less memory) or HNSW

Embedding model?
├─ text-embedding-3-small (1536) → vector(1536)
├─ text-embedding-3-large (3072) → halfvec(3072) (50% memory savings)
└─ Other models → vector(dimensions)

import { CohereClient } from 'cohere-ai';

const cohere = new CohereClient({ token: process.env.COHERE_API_KEY });

async function rerankResults(query: string, documents: string[]) {
  const response = await cohere.rerank({
    model: 'rerank-v4.0-fast',  // or 'rerank-v4.0-pro' for best quality
    query,
    documents,
    topN: 10,
  });
  return response.results;
}

// Semantic search
const { data } = await supabase.rpc('match_documents', {
  query_embedding: embedding,
  match_threshold: 0.7,
  match_count: 10
});

// Hybrid search
const { data } = await supabase.rpc('hybrid_search_fts', {
  query_embedding: embedding,
  query_text: userQuery,
  match_count: 10,
  rrf_k: 60,
  fts_language: 'simple'
});

import { sql } from 'drizzle-orm';

const results = await db.execute(sql`
  SELECT * FROM match_documents(
    ${embedding}::vector(1536),
    0.7,
    10
  )
`);