langchain4j-vector-stores-configuration

Name: langchain4j-vector-stores-configuration
Author: giuseppe-trisciuoglio

1.6kinstalls36forks

How to install langchain4j-vector-stores-configuration?

giuseppe-trisciuoglio/developer-kit/langchain4j-vector-stores-configuration

$npx -y skills add giuseppe-trisciuoglio/developer-kit --skill langchain4j-vector-stores-configuration

Installs into the current project.

Summary

Provides configuration patterns for LangChain4J vector stores in RAG applications. Use when building semantic search, integrating vector databases (PostgreSQL/pgvector, Pinecone, MongoDB, Milvus, Neo4j), implementing embedding storage/retrieval, setting up hybrid search, or optim

Files · 1

View on GitHub

SKILL.md

1	# LangChain4J Vector Stores Configuration
2
3	Configure vector stores for Retrieval-Augmented Generation applications with LangChain4J.
4
5	## Overview
6
7	LangChain4J provides a unified abstraction for vector stores (PostgreSQL/pgvector, Pinecone, MongoDB Atlas, Milvus, Neo4j) with builder-based configuration, metadata filtering, and hybrid search support.
8
9	## When to Use
10
11	- Configuring vector stores for semantic search and RAG applications
12	- Setting up embedding storage with metadata filtering and hybrid search
13	- Optimizing vector database performance for production AI workloads
14
15	## Instructions
16
17	### Set Up Basic Vector Store
18
19	Configure an embedding store for vector operations:
20
21	```java
22	@Bean
23	public EmbeddingStore<TextSegment> embeddingStore() {
24	return PgVectorEmbeddingStore.builder()
25	.host("localhost")
26	.port(5432)
27	.database("vectordb")
28	.user("username")
29	.password("password")
30	.table("embeddings")
31	.dimension(1536) // OpenAI embedding dimension
32	.createTable(true)
33	.useIndex(true)
34	.build();
35	}
36	```
37
38	### Validation Workflow
39
40	Follow this workflow to ensure correct vector store setup:
41
42	1. Configure: Build the embedding store with required dimensions and connection parameters
43	2. Test connection: Verify store connectivity with a health check before ingesting data
44	3. Validate dimensions: Confirm embedding model dimensions match store configuration
45	4. Ingest test data: Add a small batch of test documents to verify ingestion works
46	5. Run test query: Execute a sample semantic search to confirm retrieval accuracy
47	6. Proceed to production: Only after all steps pass, proceed with full data ingestion
48
49	### Configure Multiple Vector Stores
50
51	Use different stores for different use cases:
52
53	```java
54	@Configuration
55	public class MultiVectorStoreConfiguration {
56
57	@Bean
58	@Qualifier("documentsStore")
59	public EmbeddingStore<TextSegment> documentsEmbeddingStore() {
60	return PgVectorEmbeddingStore.builder()
61	.table("document_embeddings")
62	.dimension(1536)
63	.build();
64	}
65
66	@Bean
67	@Qualifier("chatHistoryStore")
68	public EmbeddingStore<TextSegment> chatHistoryEmbeddingStore() {
69	return MongoDbEmbeddingStore.builder()
70	.collectionName("chat_embeddings")
71	.build();
72	}
73	}
74	```
75
76	### Implement Document Ingestion
77
78	Use EmbeddingStoreIngestor for automated document processing:
79
80	```java
81	@Bean
82	public EmbeddingStoreIngestor embeddingStoreIngestor(
83	EmbeddingStore<TextSegment> embeddingStore,
84	EmbeddingModel embeddingModel) {
85
86	return EmbeddingStoreIngestor.builder()
87	.documentSplitter(DocumentSplitters.recursive(
88	300, // maxSegmentSizeInTokens
89	20, // maxOverlapSizeInTokens
90	new OpenAiTokenizer(GPT_3_5_TURBO)
91	))
92	.embeddingModel(embeddingModel)
93	.embeddingStore(embeddingStore)
94	.build();
95	}
96	```
97
98	### Set Up Metadata Filtering
99
100	Configure metadata-based filtering capabilities:
101
102	```java
103	// MongoDB with metadata field mapping
104	IndexMapping indexMapping = IndexMapping.builder()
105	.dimension(1536)
106	.metadataFieldNames(Set.of("category", "source", "created_date", "author"))
107	.build();
108
109	// Search with metadata filters
110	EmbeddingSearchRequest request = EmbeddingSearchRequest.builder()
111	.queryEmbedding(queryEmbedding)
112	.maxResults(10)
113	.filter(and(
114	metadataKey("category").isEqualTo("technical_docs"),
115	metadataKey("created_date").isGreaterThan(LocalDate.now().minusMonths(6))
116	))
117	.build();
118	```
119
120	### Configure Production Settings
121
122	Implement connection pooling and monitoring:
123
124	```java
125	@Bean
126	public EmbeddingStore<TextSegment> optimizedPgVectorStore() {
127	HikariConfig hikariConfig = new HikariConfig();
128	hikariConfig.setJdbcUrl("jdbc:postgresql://localhost:5432/vectordb");
129	hikariConfig.setUsername("username");
130	hikariConfig.setPassword("password");
131	hikariConfig.setMaximumPoolSize(20);
132	hikariConfig.setMinimumIdle(5);
133	hikariConfig.setConnectionTimeout(30000);
134
135	DataSource dataSource = new HikariDataSource(hikariConfig);
136
137	return PgVectorEmbeddingStore.builder()
138	.dataSource(dataSource)
139	.table("embeddings")
140	.dimension(1536)
141	.useIndex(true)
142	.build();
143	}
144	```
145
146	### Implement Health Checks
147
148	Monitor vector store connectivity:
149
150	```java
151	@Component
152	public class VectorStoreHealthIndicator implements HealthIndicator {
153
154	private final EmbeddingStore<Tex

People who liked this also viewed

6 picks

Type

langchain4j-vector-stores-configuration

1.6kinstalls36forks

View on GitHub

How to install langchain4j-vector-stores-configuration?

giuseppe-trisciuoglio/developer-kit/langchain4j-vector-stores-configuration

$npx -y skills add giuseppe-trisciuoglio/developer-kit --skill langchain4j-vector-stores-configuration

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# LangChain4J Vector Stores Configuration
2
3	Configure vector stores for Retrieval-Augmented Generation applications with LangChain4J.
4
5	## Overview
6
7	LangChain4J provides a unified abstraction for vector stores (PostgreSQL/pgvector, Pinecone, MongoDB Atlas, Milvus, Neo4j) with builder-based configuration, metadata filtering, and hybrid search support.
8
9	## When to Use
10
11	- Configuring vector stores for semantic search and RAG applications
12	- Setting up embedding storage with metadata filtering and hybrid search
13	- Optimizing vector database performance for production AI workloads
14
15	## Instructions
16
17	### Set Up Basic Vector Store
18
19	Configure an embedding store for vector operations:
20
21	```java
22	@Bean
23	public EmbeddingStore<TextSegment> embeddingStore() {
24	return PgVectorEmbeddingStore.builder()
25	.host("localhost")
26	.port(5432)
27	.database("vectordb")
28	.user("username")
29	.password("password")
30	.table("embeddings")
31	.dimension(1536) // OpenAI embedding dimension
32	.createTable(true)
33	.useIndex(true)
34	.build();
35	}
36	```
37
38	### Validation Workflow
39
40	Follow this workflow to ensure correct vector store setup:
41
42	1. Configure: Build the embedding store with required dimensions and connection parameters
43	2. Test connection: Verify store connectivity with a health check before ingesting data
44	3. Validate dimensions: Confirm embedding model dimensions match store configuration
45	4. Ingest test data: Add a small batch of test documents to verify ingestion works
46	5. Run test query: Execute a sample semantic search to confirm retrieval accuracy
47	6. Proceed to production: Only after all steps pass, proceed with full data ingestion
48
49	### Configure Multiple Vector Stores
50
51	Use different stores for different use cases:
52
53	```java
54	@Configuration
55	public class MultiVectorStoreConfiguration {
56
57	@Bean
58	@Qualifier("documentsStore")
59	public EmbeddingStore<TextSegment> documentsEmbeddingStore() {
60	return PgVectorEmbeddingStore.builder()
61	.table("document_embeddings")
62	.dimension(1536)
63	.build();
64	}
65
66	@Bean
67	@Qualifier("chatHistoryStore")
68	public EmbeddingStore<TextSegment> chatHistoryEmbeddingStore() {
69	return MongoDbEmbeddingStore.builder()
70	.collectionName("chat_embeddings")
71	.build();
72	}
73	}
74	```
75
76	### Implement Document Ingestion
77
78	Use EmbeddingStoreIngestor for automated document processing:
79
80	```java
81	@Bean
82	public EmbeddingStoreIngestor embeddingStoreIngestor(
83	EmbeddingStore<TextSegment> embeddingStore,
84	EmbeddingModel embeddingModel) {
85
86	return EmbeddingStoreIngestor.builder()
87	.documentSplitter(DocumentSplitters.recursive(
88	300, // maxSegmentSizeInTokens
89	20, // maxOverlapSizeInTokens
90	new OpenAiTokenizer(GPT_3_5_TURBO)
91	))
92	.embeddingModel(embeddingModel)
93	.embeddingStore(embeddingStore)
94	.build();
95	}
96	```
97
98	### Set Up Metadata Filtering
99
100	Configure metadata-based filtering capabilities:
101
102	```java
103	// MongoDB with metadata field mapping
104	IndexMapping indexMapping = IndexMapping.builder()
105	.dimension(1536)
106	.metadataFieldNames(Set.of("category", "source", "created_date", "author"))
107	.build();
108
109	// Search with metadata filters
110	EmbeddingSearchRequest request = EmbeddingSearchRequest.builder()
111	.queryEmbedding(queryEmbedding)
112	.maxResults(10)
113	.filter(and(
114	metadataKey("category").isEqualTo("technical_docs"),
115	metadataKey("created_date").isGreaterThan(LocalDate.now().minusMonths(6))
116	))
117	.build();
118	```
119
120	### Configure Production Settings
121
122	Implement connection pooling and monitoring:
123
124	```java
125	@Bean
126	public EmbeddingStore<TextSegment> optimizedPgVectorStore() {
127	HikariConfig hikariConfig = new HikariConfig();
128	hikariConfig.setJdbcUrl("jdbc:postgresql://localhost:5432/vectordb");
129	hikariConfig.setUsername("username");
130	hikariConfig.setPassword("password");
131	hikariConfig.setMaximumPoolSize(20);
132	hikariConfig.setMinimumIdle(5);
133	hikariConfig.setConnectionTimeout(30000);
134
135	DataSource dataSource = new HikariDataSource(hikariConfig);
136
137	return PgVectorEmbeddingStore.builder()
138	.dataSource(dataSource)
139	.table("embeddings")
140	.dimension(1536)
141	.useIndex(true)
142	.build();
143	}
144	```
145
146	### Implement Health Checks
147
148	Monitor vector store connectivity:
149
150	```java
151	@Component
152	public class VectorStoreHealthIndicator implements HealthIndicator {
153
154	private final EmbeddingStore<Tex

People who liked this also viewed

6 picks

Type