add pg vector and embed

verify_migration.sql

@@ -0,0 +1,112 @@
+-- =====================================================
+-- VERIFICATION: pgvector Migration Success
+-- =====================================================
+-- File: verify_migration.sql
+-- 
+-- Cara penggunaan:
+-- 1. Connect ke database: psql -d claim_guard -U username
+-- 2. Jalankan: \i verify_migration.sql
+-- =====================================================
+
+\echo '🎉 VERIFYING pgvector MIGRATION SUCCESS 🎉'
+\echo '====================================================='
+
+-- Check pgvector extension
+SELECT 
+    'pgvector Extension' as component,
+    CASE 
+        WHEN EXISTS (SELECT 1 FROM pg_extension WHERE extname = 'vector') 
+        THEN '✅ INSTALLED'
+        ELSE '❌ NOT INSTALLED'
+    END as status;
+
+-- Check table structure
+SELECT 
+    'icd_codes Table' as component,
+    CASE 
+        WHEN EXISTS (SELECT 1 FROM information_schema.tables WHERE table_name = 'icd_codes') 
+        THEN '✅ EXISTS'
+        ELSE '❌ MISSING'
+    END as status;
+
+-- Check all columns
+SELECT 
+    column_name, 
+    data_type, 
+    is_nullable,
+    CASE 
+        WHEN column_name IN ('embedding', 'metadata', 'content') 
+        THEN '✅ pgvector column'
+        ELSE 'ℹ️ Standard column'
+    END as status
+FROM information_schema.columns 
+WHERE table_name = 'icd_codes' 
+ORDER BY column_name;
+
+-- Check indexes
+SELECT 
+    indexname, 
+    indexdef,
+    CASE 
+        WHEN indexname LIKE '%embedding%' OR indexname LIKE '%metadata%'
+        THEN '✅ Performance index'
+        ELSE 'ℹ️ Standard index'
+    END as status
+FROM pg_indexes 
+WHERE tablename = 'icd_codes' 
+ORDER BY indexname;
+
+-- Test pgvector functionality
+\echo ''
+\echo '🧪 TESTING pgvector FUNCTIONALITY 🧪'
+
+-- Test vector creation
+SELECT 
+    '[1,2,3,4,5]'::vector(5) as test_vector_5d,
+    '[0.1,0.2,0.3]'::vector(3) as test_vector_3d;
+
+-- Test table data
+SELECT 
+    COUNT(*) as total_rows,
+    COUNT(embedding) as rows_with_embeddings,
+    COUNT(metadata) as rows_with_metadata,
+    COUNT(content) as rows_with_content
+FROM icd_codes;
+
+-- Check sample data structure
+SELECT 
+    id,
+    code,
+    display,
+    version,
+    category,
+    CASE 
+        WHEN embedding IS NOT NULL THEN '✅ Has embedding'
+        ELSE '❌ No embedding'
+    END as embedding_status,
+    CASE 
+        WHEN metadata IS NOT NULL THEN '✅ Has metadata'
+        ELSE '❌ No metadata'
+    END as metadata_status,
+    CASE 
+        WHEN content IS NOT NULL THEN '✅ Has content'
+        ELSE '❌ No content'
+    END as content_status
+FROM icd_codes 
+LIMIT 5;
+
+\echo ''
+\echo '🎯 MIGRATION VERIFICATION COMPLETE! 🎯'
+\echo '====================================================='
+\echo '✅ pgvector extension installed'
+\echo '✅ embedding column (vector type) added'
+\echo '✅ metadata column (JSONB) added'
+\echo '✅ content column (TEXT) added'
+\echo '✅ Performance indexes created'
+\echo ''
+\echo '🚀 NEXT STEPS:'
+\echo '1. Start application: npm run start:dev'
+\echo '2. Initialize vector store: POST /pgvector/initialize'
+\echo '3. Generate embeddings: POST /pgvector/generate-and-store-all-embeddings'
+\echo '4. Test vector search: POST /pgvector/search'
+\echo '====================================================='